Gender, personal values, strategies and small business performance

A Ghanaian case study

The Authors

Rosemond Boohene, New England Business School, University of New England, Armidale, Australia

Alison Sheridan, New England Business School, University of New England, Armidale, Australia

Bernice Kotey, New England Business School, University of New England, Armidale, Australia

Abstract

Purpose – While most studies concerning gender differences in small business performance have emerged from developed countries, how applicable the results are to transitional economies, where there still exist significant differences in the socialisation of men and women, is not clear. The purpose of this paper is to examine the influence of gender on the performance of small businesses in Ghana by exploring the impact of personal values on business owners' choice of strategies, specifically asking, do women and men pursue different business strategies, how do personal values influence their strategies and how do their strategies affect performance?

Design/methodology/approach – A total of 600 owner-managers of small retail shops in the Greater Accra Region were surveyed. Results were analysed using the partial least squares approach to structural equation modelling.

Findings – The results suggest that there are gender differences in personal values, which lead to different strategies adopted by women and men, which in turn influence performance; specifically Ghanaian women owner-managers are more risk-averse than Ghanaian men, and this affects their pursuit of specific functional strategies and ultimately their performance in varied ways.

Research limitations/implications – Further studies will need to replicate these findings with other types of businesses, in other locations.

Practical implications – The results suggest that closer attention should be given to developing the entrepreneurial values of women if the goal of greater economic development is to be achieved in transitional economies.

Originality/value – Limited research has been conducted on women and men small business owners in transitional economies such as Ghana. This empirical research provides important insights into this field.

Article Type:

Case study

Keyword(s):

Small enterprises; Gender; Social values; Business performance; Ghana.

Journal:

Equal Opportunities International

Volume:

Number:

Year:

2008

pp:

237-257

Copyright ©

Emerald Group Publishing Limited

ISSN:

0261-0159

Introduction

The role of small businesses in facilitating global economic development has attracted increasing attention over the past decade (Verheul and Thurik, 2001). The small business sector is seen as an important contributor in the transition to a market economy, through the processes of innovation, job creation, income generation and business dynamics (Hisrich and Ozturk, 1999). To this end, the World Bank and other international agencies have linked funding for transitional economies to policies and programmes aimed at creating stable macro-economic environments conducive to private investment, with a particular emphasis on developing the small and medium enterprise sector. However, in the majority of cases, these programmes have failed to achieve the desired outcomes because they are not tailored to the specific needs of the recipient countries (Baffoe, n.d.) In Ghana, for example, the small business sector accounted for 15.5 per cent of the labour force and 6 per cent of the gross domestic product in 1998 (Kayanula and Quartey, 2000). Despite efforts to improve this sector, the growth and performance of these enterprises in terms of their contribution to income, innovation, employment generation and equity has been disappointing (Bani, 2003).

The contribution of the small enterprise sector to overall economic development is dependent on the performance of the individual firms that comprise the sector. Performance of individual firms is in turn determined by the strategies adopted by their owner-managers. In small firms, the choice of strategies implemented reflects the owner-manager's personal goals and desires which are grounded in their personal values (Kotey and Meredith, 1997; Olson and Currie, 1992). Research has also shown that personal values, business strategy and enterprise performance are influenced by the demographic characteristics of the owner-managers, in particular, their gender (Carter et al., 1997; Verheul et al., 2002).

Most studies concerning male and female performance in small business, particularly those exploring the relationship between personal values, business strategies and performance have emerged from developed countries. It is not clear from the literature the extent to which the results from these studies apply in transitional economies, where there still exist large differences in the socialisation of men and women and their ability to access resources. In this paper, we consider the particular context of Ghana as a transitional economy, and the persistence of highly gendered opportunity structures within that economy. It is in this context that we investigate the influence of gender on the performance of small businesses by exploring the impact of personal values on business owners' choice of strategies – specifically we ask, do women and men pursue different business strategies, how do personal values influence their strategies and how do their strategies affect performance? Drawing on data collected from 600 owner-managers of small retail shops in the Greater Accra Region, our results suggest that there are significant differences between men's and women's personal values, which influence their adoption of specific business strategies, which in turn have a significant influence on performance. It seems that Ghanaian women owner-managers are more risk averse than their male counterparts, and that this affects their pursuit of specific functional strategies in varied ways. These findings suggest that raising the performance of women-owned businesses, through efforts to foster women's openness to adopting more entrepreneurial strategies, could be a critical element in the overall strategy of transitional economies to achieve sustainable growth and gender equity as well as reduce poverty (Amu, 2005; Bani, 2003).

Study context

The Ghanaian economy, as a transitional economy, is characterised by economic restructuring, involving privatisation at both macro-economic and microeconomic levels, a weak financial sector, decentralisation and deregulation of major institutions previously run by the public sector, and support for entrepreneurship and free market competition (Asfaw and Jones, 1999; Kalyuzhnova and Taylor, 2001). As in many transitional economies, the majority of urban dwellers in Ghana earn their living from the informal sector, particularly small-scale non-farming activities (Greenstreet, 1981; Republic of Ghana, n.d.). With the urban population continuously increasing, the informal sector continues to be the most important income-earning area, particularly for women (ABANTU for Development, 2004; Amu, 2005; Akuffo, 1996; Greenstreet, 1981; Tsikata, 2001). Traditionally, women in Ghana are the major source of livelihood for a majority of households although their activities are not normally included in the national accounts. Women commonly occupy key positions alongside men in the production of goods and services, and have excelled in operating small businesses (Amu, 2005). However, the majority of Ghanaian female business owners are in the trading and service sectors (ABANTU for Development, 2004; Vande-Pallen, 2006). Entrepreneurship in service and retail are viewed as part of the non-sponsored segment of the labour market, where there is an absence of government support in terms of loans or grants. This means the majority of women-owned businesses fall into the non-sponsored category of the market.

Small private enterprise does not fit well with the role expectation of Ghanaian men. Until the 1980s, men were looked down on if they engaged in small business activities; rather, they achieved social respect if employed in large multi-national firms, the public sector, managed large manufacturing concerns or were self-employed in professional services. Accordingly, women are concentrated in small-scale trading while men are involved with large-scale trading (Amu, 2005). The disparity between the sexes in labour market participation is not only due to differences in access to markets, trading contacts and transportation but also because large-scale trading requires a level of flexibility and mobility that women do not have unless freed from their domestic chores. In contrast, women's retail trade activities are carried out in close proximity to their homes, in markets or kiosks along streets or in front of homes. The majority of women in this sector are illiterate and have few assets (Amu, 2005). They lack basic skills in accounting, bookkeeping and income management and are not able to prepare the business plans and feasibility studies required by banks (Nikoi, 1993; Quainoo, 2001; Tsikata, 2001). Thus, women operating in this sector face a combination of constraints which require support in public policy rather than being abandoned to the gales of unbridled liberalisation (Vande-Pallen, 2006). A World Bank report in 1999 on Ghana's gender strategy suggested that although women are economically active in Ghana and are likely to be involved in businesses, their success is constrained by cultural, educational and economic barriers created by the acculturation and socialisation process (Bani, 2003; Quainoo, 2001).

The Ghanaian society is an interrelated social system where the extended family forms the core around which the traditional social organisation revolves (Federal Research Division, 1999). Relationships are built around family membership, inherited status and ancestral beliefs (Gyekye, 2003). Family, school and friends function as socialising agencies to ensure that societal obligations and privileges are properly carried out, although the moral and ethical instruction of children remain the responsibility of the extended family. Values are transmitted through proverbs, songs, rituals and associated rights of passage. While social behaviour and values are partially altered to fit the needs of modern life, traditional values continue to endure (Amu, 2005).

In Ghana, traditional sex-role identification and living arrangements have established clear delineations of household responsibilities (Bortey and Dodoo, n.d.). This division of labour is passed on to children within the household who learn from household chores the gender roles expected of them. When children are taught to behave in gender-specific ways, they begin to act in a manner consistent with their sex, allowing for the internalisation of gender-specific roles (Dolphyne, 2000). With no formal social security system, males in the Ghanaian society are perceived to be a form of security for their family so they have to be well socialised to perform this role (Boohene et al., 2005). They are taught to become heads of households and breadwinners charged with responsibility for the welfare of all household members including women. In the process they are given control over the resources necessary for discharging these responsibilities (Amu, 2005; Dolphyne, 2000). A male is encouraged to be aggressive, and to display “machismo”. These qualities assist in building their self-esteem and confidence (Bortey and Dodoo, n.d.). Thus, the key developmental areas in the values of boys in the Ghanaian society are assertiveness, risk taking, independence, innovativeness and achievement orientation. What the Ghanaian male achieves through intimate connection with others, he must manage as a man by disconnecting, by separating himself and asserting his right to be distinct (Bortey and Dodoo, n.d.).

Ghanaian women, on the other hand, are not expected to develop such qualities. They are not encouraged to display the same degree of assuredness in their abilities (Dolphyne, 2000). They bear the primary responsibility for child-nurturing, cooking, washing and collecting fuel-wood and water (Boohene et al., 2005; Nikoi, 1993). They are socialised to be gentle, to put the needs of others before theirs but never to be winners. Moreover, sanctions are imposed on women when they go beyond the boundaries governing the sexual division of labour (Dolphyne, 2000). They are subjected to malicious gossip and ostracised when they behave outside the economic and social norms set by the society (Greenstreet, 1981). An outcome of the different socialisation of the genders is that women may not believe in their own ability to achieve in the way that men do (Boohene et al., 2005).

In principle, the Ghanaian educational system is designed to serve girls and boys equally (Amu, 2005; Greenstreet, 1981). The curriculum, however, divides the practical subjects into those suitable for girls and those suitable for boys. During adolescence, boys and girls are presented with different expectations of their potential based on their gender (Greenstreet, 1981). In addition, women are increasingly channelled into secondary jobs, mainly as a result of limited access to education and training (Akuffo, 1996; Greenstreet, 1981; Nikoi, 1993). It is explained that in traditional societies, the major role of a woman is to ensure the continuity of the family lineage. She is therefore, expected to marry soon after puberty and does not need formal education to perform this role. Moreover, a man is expected to provide for his wife and children. Since education is a means to highly paid jobs in the formal sector, it is considered more important for boys to have formal education to enable execution of their roles as breadwinners in the family (Dolphyne, 2000; Tsikata, 2001).

In addition, the education of Ghanaian females is hindered by the overall cost of education, customary fostering, puberty rites, early marriages, “trokosi” (female ritual slavery) and the perception of parents that educating girls is not as rewarding as educating boys (Amu, 2005; Republic of Ghana, 2004). Thus, most girls do not continue their education after high school (Federal Research Division, 1999). Others do not even complete an elementary level of education because they are withdrawn from school to help with household chores and family businesses (Republic of Ghana, 2004). This social attitude towards female education is linked to the conviction that a woman will be supported by her husband.

Despite school enrolment and retention rates in Ghana have increasing for both boys and girls in recent years, the gender gap in enrolment and retention still persists, widening at higher grades (Republic of Ghana, 2004). The 2004 progress report on Ghana's millennium development goals revealed various inequalities between the genders. At the primary level, the gender parity index was 93 females to 100 males between 2003 and 2004. However, the differences widened at the junior secondary school level to 88 for females and 100 for males. In a related report, the ratio of male to female enrolment in Ghanaian public universities was presented as 70-30, for 2002 (Institute of Statistical Social and Economic Research, 2003).

Whilst modern lifestyles and Western influences have reduced the disparities in nurturing between the genders, particularly among urban dwellers, Ghanaians are still bound through the kinship system to traditional social values that require them to assume the responsibilities that such associations entail (Bortey and Dodoo, n.d.; Federal Research Division, 1999). Girls and women are pushed into female stereotyped careers such as teaching, tailoring, secretarial work and nursing, and are discouraged from entering professions in agriculture, forestry, fishing, “hard” sciences, engineering and accounting. As a result, they have less opportunity to acquire the experiences necessary to operate businesses successfully. Although Ghana has not experienced anything close to the women's movement that occurred in Western countries in the 1960s, gender inequities in the workforce bear some resemblance to that observed in Western countries (Boohene et al., 2005). Changes associated with the push for greater equity in access to resources among the sexes have been slow because women still face discrimination in both the formal and informal sectors of the economy (Tsikata, 2001). Beyond these factors is the common perception that women's businesses should not be as large or successful as men's (The World Bank Group, 1999). Consequently, studies in the small business literature have indicated that, in order to improve the performance of the small business sector in general and the status of women, in particular, greater attention should be paid to values, family situation, resources, managerial experience, business skills, business strategies and educational levels all of which are influenced by cultural and socialisation practices (Bani, 2003; Quainoo, 2001).

From the foregoing discussions, it is clear that women in Ghana must overcome structural barriers in the creation and management of their businesses if they are to be successful. Clearly, the economic, social and cultural environment within which men and women operate their businesses has an important influence on their personal values, strategic options at their disposal and ultimately the performance of their businesses.

Research framework

In light of the role, small businesses play in transitional economies, we thought it timely to explore whether the differences in performance, business strategies and personal values between the genders in Ghana. To what extent do gender differences in personal values influence business strategies which ultimately influence business performance? If so, what are the implications for policy-making designed to promote the small business sector? In addressing these questions, we examine the effects of gender on personal values, strategy and performance.

Personal values, strategy and performance

Strategy is recognised as a key aspect of management in large organisations (Gibcus and Kemp, 2003). As such large firms often have written or formal statements of their strategies. In contrast, strategies are less formal in small firms and originate from the owner-manager who is the key decision-maker (McCarthy, 2003). Since small firms do not normally have written statements of strategy, their strategies are inferred from evolving patterns of owner-manager behaviour and resource allocation (Schindehutte and Morris, 2001). Strategy in small firms may or may not be intentional and generally emerges through a series of incremental adjustments to the opportunities and threats confronted by the firm over-time. Strategies are based on the owner-manager's subjective interpretations of and reactions to the day-to-day situations confronted by the firm (D' Amboise, 1993; Williams and Tse, 1995). In small firms, then, business strategies are deduced from activities in the major functional areas of finance, human resource management (HRM), marketing, production and operations and information and communications technology. These activities act as a guide to the overall business strategy realized by the firm (Galbraith and Schendel, 1983).

The performance of small firms has been found to be contingent upon the strategies they use (Covin, 1991). That is, the ability to formulate and implement effective strategies has a major influence on the survival or failure of small businesses. For instance, Rue and Ibrahim (1998) found positive associations between strategy and firm performance. The literature presents numerous types of strategies and their impact on performance. Popular among these are Miles and Snow's (1978) four strategy-types and Porter's (1980) five competitive strategies. In the small business literature, strategy-types have been presented on a continuum ranging from entrepreneurial-conservative or proactive-reactive (Covin, 1991). The literature on strategic planning and firm performance suggests that proactive firms invest time in developing strategies and exhibit higher performance levels than less proactive firms which do not develop strategies, but react to events in their environment (Covin et al., 1990). Other reported differences between proactive and reactive strategies include the level of creativity, risk taking, innovation and planning (Lumpkin and Dess, 1997).

Fagenson (1993) argued that values exert a strong influence on an individual's vocational behaviour. The specific behaviour of interest in this paper is the behaviour of owner-managers in small firms. Accordingly, a key factor in the creation and implementation of strategy in small businesses is the determining influence of the owner-manager's values (Olson and Currie, 1992). Owner-managers' values determine the goals pursued by the business and influence their management style, problem-solving and decision-making behaviour as well as the structure of the business and strategies employed (Olson and Currie, 1992; Verheul et al., 2002). Thompson and Strickland (1984, p. 49) noted that:

Strategy formulation is rarely so objective an analytical process that the personal values and aspirations of managers are excluded… There is a natural human tendency for managers to draw upon their own personal values, background experiences, preferences, and ambitions when choosing among alternative strategies and when interpreting the strategic plan.

The personal values associated with proactive strategies, often referred to as entrepreneurial values, cover creativity, risk taking, innovation, achievement-oriented, ambition and independence (Hodgetts and Kuratko, 2001; Kickul and Gundry, 2002), are all terms highly associated with masculinity (Schein, 1973). In contrast, values such as equity, affection, social protection and compassion are described as conservative and associated with reactive type strategies (Kotey, 1994) and commonly associated with femininity (Schein, 1973). The literature indicates that owner-managers with entrepreneurial values who adopt proactive strategies exhibit stronger performance than those with less entrepreneurial values who are reactive in strategic orientation (Blackman, 2003).

Carter et al. (2001) and Hisrich et al. (1997) argued that if the execution of a firm's strategy is influenced by the owner-manager's values, experiences, motivations and belief structures, then male and female owner-managers should differ in the strategies they adopt in operating their businesses. This is because differences in the socialisation of men and women suggest they will exhibit different personal values (Fagenson, 1993). We explore the association between gender and the three variables (personal values, strategies and business performance) in the next section.

Gender, personal values, strategies and performance

A generally accepted premise in the small business literature is that female owner-managers behave differently from male owner-managers and that several factors unique to women influence the strategies they adopt and ultimately performance of their business.

Downing (1990) suggested that personal values and objectives of women differ to those of men. This is because men and women undergo different socialisation processes, where each sex acquires specific sex-linked personal values and learns sex-appropriate behaviours (Bird and Brush, 2002). Men have been traditionally socialised to be more material and achievement-oriented and pleasure seeking than women. This is demonstrated in their value for innovation, risk taking, independence, capability, courage, ambition, logic, a comfortable life, a sense of accomplishment, freedom and family security (Fagenson, 1993; Rokeach, 1973; Verheul et al., 2002). The majority of these values are consistent with those presented in the literature as entrepreneurial. On the other hand, women have been socialised to be care givers, oriented towards religious values with less emphasises on the things valued by men (Boohene et al., 2005). As such, the following hypothesis for testing is proposed:

H1a. Ghanaian female owner-managers will exhibit lower levels of entrepreneurial values compared to male owner-managers.

Gendered social relations shape individual decisions that affect opportunities and at the same time influence the accessibility of these opportunities. Thus, Carter et al. (1997) proposed that as women and men are socialised to have diverse values, then they will pursue different objectives and ultimately adopt different strategies in managing their businesses. Kalleberg and Leicht (1991) argued that women are usually thought to be less likely than men to engage in entrepreneurial behaviour regarded as essential to business growth and development. Compared to men, women have been found to be more cautious (Jianakoplos and Bernasek, 1998), less confident, less aggressive, easier to persuade, and have lower problem-solving abilities when making decisions under risk conditions (Johnson and Powell, 1994; Sonfield et al., 2001). Powell and Ansic (1997), for instance, noted that women have been stereotyped as being conservative and risk averse, whereas males are viewed as more entrepreneurial and risk taking.

If women are more cautious and risk averse than men, and if risk aversion is characteristic of reactive strategy, then it could be expected that compared to men, women will be less proactive in their strategic orientation. In their study on strategies adopted by female entrepreneurs, Olson and Currie (1992) and Verheul et al. (2002) reported that female entrepreneurs adopt a more defensive strategy compared to male entrepreneurs who were associated with aggressive strategies. Accordingly, the hypothesis below is proposed for testing:

H1b. Ghanaian female owner-managers will be less proactive with their choice of strategies than their male counterparts.

Gender differences in small business performance were largely unaddressed by social scientists for decades (Greene et al., 2003). Most studies either did not consider gender as a variable of interest or excluded female subjects from their design (Du Reitz and Henrekson, 2000). However, more current studies have shown stereotypical differences in performance between the sexes (Litz and Folker, 2002). Based on H1b, if women owner-managers are less proactive than male owner-managers, then it is expected that they will exhibit poorer performance. The following hypothesis is proposed for testing.

H1c. Ghanaian female owner-managers' businesses will exhibit lower performance levels compared to their male counterparts.

From the forgoing discussion, if personal values influence business strategies and if the gender of the owner-manager influences these factors, then the gender of the owner-manager will influence the relationships between personal values, business strategies and ultimately firm performance. The following hypothesis is proposed to examine the indirect effect of gender of owner-managers on their business performance, based on their personal values and strategies:

H2. Ghanaian female owner-managers will exhibit lower levels of entrepreneurial values, adopt less proactive strategies and so will exhibit poorer performance compared to their male counterparts.

Sample and variables

In order to reduce the confounding effects of any disparities in firm characteristics and background variables such as industry type, technology, opportunity, location, sectoral activities and socio-political variables, a single industry spatially limited to one area, the Accra Metropolis (Accra Central), Ghana, was selected. Furthermore, it was decided that this study would maintain some degree of homogeneity in the type of businesses examined by focusing only on general merchants in the retail sector in Accra, as this would reduce the variance error associated with inter-industry variations and establish similar standards to allow comparability. Thus, the retail sector was chosen for three reasons. Firstly, an important research gap in the literature could be filled. Although small retail businesses form a large part of the small business sector in Ghana, there is little research on activities in the sector and specifically on the factors that influence the performance of firms in the sector. Reference to the small business sector by researchers and the government typically relates to small-scale manufacturing businesses.

Secondly, the small business sector in the Accra metropolis in Ghana is dominated by wholesale and retail trade (30.4 per cent) followed by manufacturing (16.7 per cent) (Republic of Ghana, n.d.). Furthermore, a higher proportion of females (39 per cent) are in wholesale and retail trade compared to males (22.2 per cent) (Republic of Ghana, n.d.). Thus, both women and men choose to operate in this sector, providing a sufficient number of female-owned businesses to compare with those of males. The limited number of women-owned firms in small manufacturing (19.5 per cent) precludes a meaningful comparison. Finally, the study was confined to a single sector because managerial variables differ greatly from one sector to another (Chaganti et al., 1989; Rosa et al., 1994).

To be included in the study, owner-managers and their small retail firms had to meet three criteria: firstly, the business had to be operating in the Accra Metropolis; secondly, it had to be classified as a small business (a business with no more than 30 employees); and finally, it had to be privately owned. As it was impossible to generate an accurate list of registered retailers in the study area, the suggestion of Bartholomew and Smith (2006) that social networks help increase the response rates from chief executive officers of small firms was employed. Thus, the Ghana Union of Traders (GUTA), an umbrella body formed to coordinate the activities of retailers in the Accra metropolitan area, was approached for information on the sector. Members of the association therefore formed the population and the bulk of respondents for the study as almost all retailers in the study area are members of this association.

Area sampling, a probability sampling method was utilised for the quantitative data collection in this study (Kumar et al., 2002). This was the most suitable sampling procedure based on financial and time constraints and on the research objectives. In addition, there was no up-to-date list of the retailers in the Accra metropolis, and most retailers are clustered in the central business district. Furthermore, area sampling is common to research in developing countries where adequate population lists are rare (FAO, n.d.). Hussey and Hussey (1997) also stated that this sampling procedure is particularly useful for face-to-face interviews where, for time or economic reasons, it is necessary to reduce the physical area covered.

Clegg (1990) suggested that, in deciding a sample size the following points should be considered: the kind of statistical analysis planned, the expected variability within the samples based on experience, and the traditions of the particular research area regarding appropriate sample size. Based on a table provided by Krejcie and Morgan (1970) on the adequacy of a sample size from a given population, the minimum sample size of a population of about 5,000 retailers in the Accra metropolis requires a sample of about 357 retailers. However, the sampling procedure produced about 674 retailers, which is a sufficient sample size. The fieldwork for this study was conducted during the period of April–October 2005. Prior to the interviews, GUTA members were informed as to when the survey will commence, and were asked to cooperate with the interviewers during the survey period. Research assistants in addition to the researcher were dispatched in pairs to the selected areas in the Accra metropolis. Respondents were asked questions relating to their values, strategic orientations and firm performance. The face-to-face interviews yielded 674 responses, out of which only 600 were useable. The remaining 74 were incomplete, with about half the responses missing.

Measurement of variables

The latent variables (constructs) used in this study were measured using multi-item observed variables (measures). For most constructs, the measures were generated from previous research and modified to fit the current research context.

The gender of the owner-manager or major decision-maker was measured on a categorical scale using 0 for men and 1 for women. Kotey's (1999) performance measures were modified to include other performance measures adopted in the literature. Kotey (1999) argued that because research and government interest in small firms are associated with macro-economic objectives such as job creation, innovation and increased productivity, enterprise performance should be measured in terms of respondents' desire to meet these and other business goals (i.e. the level of importance respondents attach to these goals) and their assessment of the extent to which these objectives were met by their firms (i.e. respondents' satisfaction with achievement of goals).

Respondents were first asked to indicate the importance they attach to ten performance goals on a six-point Likert scale, ranging from 5 = extremely important to 0 = not at all important. Secondly, the respondents were asked to indicate on a seven-point Likert scale the extent to which they are satisfied with the firm's performance on the same measures with 7 being highly satisfied and 1 highly dissatisfied. Ratings of importance were multiplied by the levels of satisfaction to compute a performance index for each firm (Miles et al., 2000). Dess and Robinson (1984) argued that performance measured in this way has been found to have high validity and reliability.

Hulland (1999) suggests that when a construct is conceptualized as multi-dimensional, it is appropriate to separately present each dimension in the statistical model. Following this suggestion, respondents were provided a list of strategic variables in the various functional areas generated from the literature. Covin and Slevin's (1989) strategy instrument was modified to include activities in the area of information technology (IT) and to reflect the unique characteristics of the sample studied. Thus, strategy was measured as comprising activities in the functional areas of planning, finances, HRM, operations, marketing and information and communications technology practices. Respondents were asked to rate the extent to which they perform the various activities in these functional areas on a six-point Likert scale, ranging from always = 5 to never = 0.

A six-point Likert scale ranging from 0 = not at all important to 5 = extremely important was used to measure personal values. Respondents were asked to indicate the level of importance they attach to each personal value measure in the instrument. Personal values measures such as risk taking, innovation, ambition, achievement, imagination, logic, responsibility, aggressiveness, self-confidence, a need for achievement, social recognition, broadmindedness, capability and courageousness were labelled as entrepreneurial (Begley and Boyd, 1986; England, 1967; Rokeach, 1973; Verheul et al., 2002).

Analytical techniques

The partial least squares (PLS) approach to structural equation modelling (using PLS graph version 3) was used in this study. PLS is a component-based technique that simultaneously examines theory (structural model) and measures (measurement model). The advantage of using this technique is its ability to handle multiple exogenous and endogenous constructs at the same time. The model is also able to deal with multi-collinearity among endogenous constructs and create latent constructs scores directly from the cross-products of the multi-item measures. Although PLS simultaneously estimates parameters for the measurement and structural models, the two models are analysed and interpreted separately in this study (Hulland, 1999).

Analysis and results

Descriptive statistics

Of the 600 responses used in the analyses 50.6 per cent are from males and 49.3 per cent from females. The majority of the respondents (69.8 per cent) were married, 30.2 per cent had never been married, separated or divorced. Marital status did not differ greatly between the sexes – 70.7 per cent of men and 68.9 per cent of women were married and 29.2 per cent and 31 per cent, respectively, of men and women respondents were single. The most common age category of respondents (20.5 per cent) was the 30-34 age group. The chi-square test did not show any significant differences in age between men and women owner-managers (χ² = 14.06; p>0.05).

The measurement model

The measurement model evaluates the relationships between constructs and their measures by assessing the reliability and validity of the instruments used. We examined the individual item reliability by the loadings of the measures to their respective constructs (see Table I). The composite reliability for the constructs range from 0.769 to 0.915, all higher than the 0.70 level suggested by Nunnally (1978).

Convergent validity was used to measure the amount of variance that the latent variables captured from its indicators relative to the amount due to measurement error. This was assessed by using the average variance extracted (AVE). AVE measures of 0.5 or more demonstrate adequate convergent validity (Chin, 1998). Magner et al. (1996), however, argued that AVEs of 0.4 are also indicative of adequate validity. As shown in Table I, the AVEs of all the constructs were 0.4 and above providing the evidence of adequate convergent validity.

The constructs for discriminate validity (the ability of some measure to have a low correlation with measures of dissimilar concepts) were evaluated by comparing the square roots of AVE to the correlation between constructs. This can be seen in the correlation matrix in Table II, with correlations among constructs in the off-diagonal and the square root of AVE in the diagonal. The diagonal elements are all greater than their respective off-diagonal elements, indicating adequate discriminant validity. Therefore, the above analysis demonstrates that the measurement model is reliable and valid.

Discussion of structural model

The structural model was used to test the hypothesized relationships between the theoretical constructs. A summary of the path coefficients and their associated t-values are presented in Table III, which shows the direct path relationships between constructs in the theoretical model including their R ²s. Table IV presents the indirect path relationships using Sobel's test for indirect effects (Preacher and Leonardelli, n.d.).

In H1a, gender was predicted to influence the personal values of owner-managers. The path coefficient in the structural model was statistically significant and negative confirming the position that Ghanaian female owner-managers adopt less entrepreneurial values compared to male owner-managers. The results provide empirical validation for the propositions of Chaganti (1986), Kalleberg and Leicht (1991) and Olson and Currie (1992) who revealed that women entrepreneurs generally possess “shared values” for business and tend to have an orientation towards conservatism. This was also expected for the Ghanaian society where compared to men, women are socialised to be less assertive, dependent, less achievement-oriented, ambitious and risk averse (Boohene et al., 2005).

H1b predicted a relationship between gender and business strategies; specifically that Ghanaian female owner-managers would be less proactive in the areas of strategic planning, financial strategy, HRM strategy, marketing strategy and information and communications strategy than Ghanaian male owner-managers. The structural path coefficients between gender and most of the strategy items were significant (see Table III), with the exception of strategic planning and operations strategy, providing partial support for the hypothesis. The results indicate that while men and women owner-managers adopt similar planning and operations strategies, women are less proactive with their financial, HRM, communications and IT strategies, but not marketing strategies where women are more proactive than men. The results concur with those for the developed economies (Olson and Currie, 1992).

H1c predicted a relationship between gender and performance; specifically that Ghanaian female owner-managers would exhibit lower performance levels than their male counterparts. Our results weakly supported H1c. This means that the gender of the owner-manager has a weak direct influence on performance.

Testing of H2 revealed that women owner-managers in this study exhibit lower entrepreneurial values and therefore pursue less proactive strategies than the men (Verheul et al., 2002). If entrepreneurial values are associated with proactive strategies, women would be expected to adopt less proactive strategies by reason of their lower ratings on entrepreneurial values. Fewer women with entrepreneurial values will limit the pursuit of proactive strategies among female owner-managers. The results indicate that while Ghanaian male and female owner-managers adopt similar (albeit informal) approaches to strategic planning (developing plans, setting goals, comparing actual performance with business goals and benchmarks) and to operations management (employee compensation, stock and asset management), Ghanaian female owner-managers rate lower than their male counterparts in the functional areas of financial management, HRM, and information and communications technology management. These results are consistent with those of Brush and Hisrich (2000) and Chaganti (1986), who suggested that the business strategies of women owner-managers differ from those of men owner-managers. Researchers also note that women tend to be more defensive in their strategic orientation than men, who usually adopt aggressive strategies (Olson and Currie, 1992; Verheul et al., 2002). While these studies examined differences in the broad strategic orientations between men and women owner-managers, this research goes a step further by investigating the specific functional areas in which women owner-managers are less effective than men.

For practices in specific functional areas, the findings of the present study are contrary to those of Brush and Hisrich (2000), who found that the women owner-managers in their sample of firms in the USA rated themselves higher on HRM strategies than men. Ghanaian female owner-mangers had lower ratings on HRM because the majority did not have employees and therefore were not especially concerned with HRM issues. Further, the conclusion that female owner-managers are less proactive in financial strategy and information and communications technology concurs with the outcomes of studies in the Western literature which found that women-owned businesses adopt a cautious and risk averse financial strategy (Chaganti, 1986), and are less proficient and unsure of whether or not to employ information and communications technology strategies in their businesses. Thus, the financial monitoring in women-owned businesses tends to be weak or non-existent (Igbaria et al., 1998). Whilst a cautious or conservative approach to business operation may have merits in terms of sustainability, businesses that merely survive add little to economic development (Acs and Storey, 2004). Women owner-managers, like their male counterparts, should be encouraged to realize their full potential through training and mentoring programmes that assist with management of their businesses and encourage pursuit of growth goals. Further, programmes that encourage entrepreneurial values among the female population should also encourage pursuit of proactive strategies, and growth and profit goals.

We further tested H2 by examining whether gender has an indirect effect on performance through business strategies. Partial support for the relationship between gender, personal values, business strategy and firm performance was obtained. The path coefficients between gender and operations strategy and marketing were not statistically significant (Table IV). The results for this hypothesis reveal that men are able to achieve a competitive advantage over women by reason of their strengths in planning, financial strategies, HRM strategies and control of operations and in IT strategies. Any unique strengths that men may have over women in the other strategic areas of operation are easily copied so that women are able to keep abreast with men in these areas. The geographical clustering of retail firms in Ghana makes it easy for operators to imitate each others' strategies particular where the barriers to imitation are low. The findings refute those of Saffu and Manu (2004) who found no relationship between strategic planning and performance in an unstable economic environment like Ghana where firms are challenged by credit constraints, increase oil prices, unfavourable fiscal and monetary policies and fluctuating exchange rates. In contrast, our findings indicate that attention to strategic planning in the face of these difficulties does enhance performance.

The findings emphasise the strategic areas where Ghanaian men have greater strengths than their female counterparts and indicate that pursuit of these strengths is fuelled by the entrepreneurial value orientations of Ghanaian men. While in general men owner-managers do not have a significant advantage over women in operations strategy, male entrepreneurial owner-managers are able to achieve higher performance than their female counterparts because they implement more proactive operations strategies. This result is consistent with the higher growth-oriented performance goals of male owner-managers compared to females. Entrepreneurial men pursuing growth-oriented performance goals give more attention to managing larger volumes of stock and perhaps more expensive equipment than their female counterparts who are likely to be more cautious in pursuing growth goals and therefore have fewer assets to manage. From a policy-making perspective, these findings suggest that programmes that foster more attention by Ghanaian women to strategic planning, financial, HRM and operations strategies should equip them with the skills required to experience higher levels of performance in their businesses should be followed.

The relationships between gender, personal values, business strategies and enterprise performance discussed above are depicted in Figure 1.

Limitations

Although the study provides useful insights by demonstrating the importance of gender in stimulating owner-manager's personal values and enhancing performance in a developing country context, the results need to be interpreted subject to the usual limitation of survey research. In particular, the focus only on the relationships within the retail sector and in only one locality (though it improved the internal validity of the study) limits the extent to which the results could be generalized across other sectors and localities.

We acknowledge that this study has not considered the factors leading to the differences in personal values reported by the women and men surveyed. While we recognise that this is an important consideration, it is beyond the scope of this paper to explore all the factors impacting on personal values. In Boohene (2006) the broader factors – for example, education, access to resources, previous work experience – impacting on personal values are explored in detail. From a policy-making perspective, these factors would need to be considered if the goal of enhancing women's openness to more entrepreneurial values is to be achieved.

Conclusions

In light of the relative dearth of research into small business performance in transitional economies, this study aimed to provide a greater understanding of the impact of gender on personal values, business strategies and firm performance and to explain differences in performance between the women and men by the indirect effect of personal values and business strategies. From the data collected from 600 owner-managers of small retail shops in the Greater Accra Region of Ghana, our results suggest that there are gender differences in personal values, which lead to different strategies adopted by women and men, which in turn influence performance.

This study confirmed the impact of gender differences on firm performance. It revealed that Ghanaian women owner-managers are more risk averse than their male counterparts, and that this affects their pursuit of specific functional strategies – strategic planning, financial, HRM and operations strategies – in varied ways. These differences are linked to the acculturation and socialisation processes within Ghanaian society. The findings thus lend credence to the suggestion that gender differences in performance arise from institutionalised social structures that influence the perceptions, attitudes and behaviours of society's members, including women (Shaw et al., 2006). The barriers facing women owner-managers are not just economic, they are also socio-cultural. Policy makers should take into account the interactions among business performance, owner-manager characteristics and social structures in formulating policies and programmes aimed at alleviating the imbalance between the sexes in access to resources.

Figure 1.Relationships between gender, personal values, business strategies and firm performance

Table I.Reliability and convergent validity (AVE) for overall measurement model

Table II.Discriminant validity for overall measurement model

Table III.Gender, personal values, business strategies and performance

Table IV.Sobel's test of indirect effects

References

ABANTU for Development (2004), The Women's Manifesto for Ghana, Cobert Impressions, Accra, .