REASONS FOR SMALL TO MEDIUM ENTERPRISE STARTUPS: A CORRESPONDENCE ANALYSIS
Geoffrey N. Soutar and Leonie V. Still Graduate School of Management University of Western Australia
SUMMARY
Small business owners/operators commence business for a number of reasons. Previous research suggests that women want autonomy, independence and flexibility, an opportunity to be more creative, an opportunity to earn more money, and are tired of working for someone else. Men want to earn more money and have an opportunity to pass something on to their family. These findings have been consistent across a number of studies, in a variety of contexts. The present study attempted to improve our understanding of the motivations for entry by analysing the startup motivations of a sample of small business operators in Western Australia using correspondence analysis, which enabled an examination of the relationships between the various startup motivations and a clustering of respondents into groups with similar startup motivations.
Differences between the groups’
backgrounds (eg gender, age, educational background) were also examined and the results obtained and their implications are discussed.
INTRODUCTION
It is clear that people start their own businesses for a variety of reasons and, for small businesses, the lack of separation between ownership and management allows the owners’ goals to become the organisation’s goals [18]. Consequently, an understanding of the reasons why people start small businesses would give insight into the nature of small businesses and the differences between big and small businesses. The present paper provides some evidence about the reasons for small business startup by examining responses to a large-scale survey of Australian small business owners and by examining the patterns of these reasons across the sample.
Much previous research in this area has tended to concentrate on the reasons why men commence small to medium sized enterprises [4]. This is because the advent of women into small business has been a fairly recent social and economic phenomenon [7, 21, 23, 24, 25]. Hence, the findings of prior research have tended to be described in economic rationalist terms, with the reasons for start-up being a desire to make more money, or to
build wealth and social position for the family. Buttner & Moore [4] describe these reasons as “pull’ factors, a feature verified by other research [e.g. 1, 20]. Other researchers have suggested that there are both economic and lifestyle reasons for starting a small business and that, for many, the significant freedom of “being your own boss’ has been a major objective [15]. Hence, a desire to work independently and to have greater control over one’s work becomes more important for such people.
Research into the motivations behind SME start-ups by women reveals some different objectives. While a “desire to be your own boss’ is also paramount among women [24, 25] and there is no difference between genders in regards to personal goals, such as independence, achievement and economic necessity [11], women rate “push’ factors more highly than men [4, 5, 6]. Thus, limited advancement opportunities seem to be more important reasons for women starting their own business.
Borzi [2, p 5], in her analysis of the gender finance gap in relation to start-ups, describes other important reasons for women: namely, the pursuit of more social goals rather than profit and growth, noting: Women typically wish to make a living, whilst men will be looking for high growth and economic wealth. This is also reflected in their goals, which for men are purely economic. Women tend to care more about wider social and economic goals, which is clear in their individual choice of goal measurements. For men this is profit. For women the measurement criteria for their goals is dominated by customer satisfaction, being the “best’, contributing to society, the environment, and so on. This latter analysis is confirmed by late 1990s research. For instance, Still & Timms [25], in their comparative gender study of small business owners, found that men nominated the capacity to earn more money, the ability to make a profit, and the capacity to pass something on to the family as the most important factors in starting their businesses. Their reasons seemed to be primarily financial. Contrariwise, women nominated the opportunity to be more creative, identifying a gap in the market, meeting a particular service or need, and the opportunity to be home with children or dependent adults as the most important factors to them. While they also acknowledged the importance of being their own boss (autonomy), and the ability to make money, they were more interested than men in providing a service and seeking independence and self-development.
Similarly, Mallette and
McGuinness [16], in their large-scale study of Canadian entrepreneurs, found that men and women emphasised somewhat different motivations. Women rated autonomy higher to gain the greater personal freedom and independence it implies, a result echoing other research where women have indicated that they sought flexibility to balance work and family, while also wanting their work to be challenging and self-fulfilling [4, 24]. On the
other hand men gave higher ratings to improved life than women.
For men this meant an improved financial
situation and a better family life (from a status perspective). In other words, men put more emphasis on economic goals than women, while women often indicate a willingness to sacrifice some economic performance in favour of social goals such as an increase in customer satisfaction [3, 7].
Thus, many of the stated reasons for starting a small business can be classified as both financial and nonfinancial, depending on the perspective and background of the small business owner. While profits and growth are important to all SME owners, other non-financial elements may be the major motivators for a significant number [14]. Indeed, each small business owner may have a unique set of reasons for starting their business, although it is likely that there will be some commonality, at least at a subgroup level.
Accordingly, this paper sets out to explore this issue by examining a sample of male and female small business owners in Western Australia, one of the major states for small business employment in Australia. The intention is to analyse startup motivation data using correspondence analysis to enable an examination of the relationships between the various startup motivations and a clustering of respondents into groups with similar startup motivation patterns.
Although startup motivations have been analysed previously, as the prior discussion
reveals, this is the first piece of research to use correspondence analysis to be more definitive in relation to motivations and respondents’ backgrounds.
The results should shed substantive light on whether or not
respondents’ backgrounds and gender are important in this respect, thereby assisting researchers and policy makers alike in this important element of SME development.
THE PRESENT STUDY
The Research Approach
A total of 1000 small business proprietors in Western Australia were surveyed through the use of a mail questionnaire; the names and addresses being obtained from the “Yellow Pages on ROM.” A total of 50 questionnaires were returned by people who were no longer in business or who did not meet the small business criterion used in the study. Useable responses were obtained from 392 small business proprietors, giving an
effective response rate of 39%, which compares well with response rates for similar surveys [17, 22, 23]. A total of 378 respondents provided all of the information needed for the present study and all such respondents were included in the analysis.
Sample Profile
Seventy-nine per cent of the respondents were males and 19% were females. Most respondents (70%) were aged between 30 and 50 years, a common age distribution found in similar studies of small business proprietors [eg 7]. On average respondents had been in business between five to seven years, although this varied from under one year (8%) to more than ten years (30%). Most respondents had started the businesses themselves (39%) or with another person (28%). Approximately one-fifth of the sample was sole traders, while another 31% were in a partnership arrangement. Almost 45% operated a proprietary company. Some 27% of the respondents operated a home-based business.
Data Analysis
It was decided to use correspondence analysis to examine the “startup” data as this statistical procedure is useful when data are categorical, as in the present “reason or not a reason to start a small business” case [8, 9, 12, 26]. Correspondence analysis, which is a form of principal components analysis for nominal data, can be interpreted in a variety of ways. Initially the spatial variation in each row or column, which is termed “inertia,” can be examined.
The eigenvalues associated with the solution show the inertia explained by the principal axes and the sum of the eigenvalues shows the total inertia explained (which is an analog to the variation explained in traditional multivariate analysis). The inertia along each axis (or dimension) can be broken down into the proportions explained by the various row (or column) points. These proportions are termed “absolute contributions” and can be used in interpreting the various axes or dimensions. The relative contributions to inertia can be shown to be the relevant squared correlations and the sum of these correlations across the dimensions is used to decide the “quality” of the results for a particular row or column variable [12]. High “quality” suggests the results obtained provide a good representation of that aspect of the data, giving insight into the number of dimensions that should be retained.
Correspondence analysis also provides co-ordinate values for each row and column point, which allows the rows and the columns of a data matrix to be “mapped.” Thus, in the present study, both reason for startup and respondents can be “mapped.” These pieces of information were both used in the subsequent analysis, although respondents’ scores were of more interest as they were used to find people’s “reason to start patterns.” In the second phase of the analysis, respondents’ “scores” were used to develop a “startup reason profile” for each person.
Following the procedure suggested by Green, Schaffer and Patterson [8] and Soutar, and McNeil [26] these “profiles” were cluster analysed. In the present study, Howard and Harris’s [13] K-means procedure was used to find if there were groups with different reasons for starting a small business as this procedure has been found to be useful in a variety of situations [26]. Since cluster analysis produces a nominal dependent variable (group membership), discriminant analysis was then used to identify background differences between the groups. The results obtained from these various analyses are outlined in the next section.
THE RESULTS OBTAINED Descriptive Statistics
As can be seen from Table 1, respondents had both economic and lifestyle reasons for starting their small business. The most common startup reason was to obtain greater independence (55%), but making a profit (48%), earning a living (44%) and earning more money (42%) were also common reasons for respondents starting their small business. Few respondents (6%) started their business because they could not find another job.
While these frequencies provide useful initial information, respondents’ patterns are of more interest when trying to understand their startup decisions. As already noted, correspondence analysis is an appropriate method to analyse such data as it allows categorical data (such as the yes/no responses to the reasons for starting a small business obtained in the present study) to be analysed in a multivariate way.
The statistical aspects of
correspondence analysis are well discussed by Greenacre [9] and Hoffman and Franke [12] and were outlined briefly in an earlier part of the present paper and the analysis was undertaken using Greenacre’s SIMCA program [10].
Table 1: Reasons for Startup Reason for startup
Percentage for whom this was true *
Greater independence and flexibility
55
To make a profit
48
To earn a living
44
To earn more money than as employee
42
Tired of working for some one else
40
More opportunity to be creative
34
Saw a gap in the market
22
To meet a particular need in my area
17
To be home with the family
9
Could not find a job
6
Other
5
* Adds up to more than one as multiple choices were allowed
After “doubling the matrix” (Hoffman and Franke [12 p. 221]) to obtain a 378 (respondents) x 20 (“doubled”) startup reason matrix, a six dimensional solution was found to be appropriate as it ensured a “high quality” result that represented respondents’ patterns well and explained 68% of the inertia in the data. As was noted earlier, correspondence analysis provides coordinate values for both the rows (respondents) and columns (reasons for startup) of a data matrix. Initially the relationships between the reasons for starting a small business were examined.
Since, as has been noted, the data suggested a six dimensional solution, the relationships cannot be easily examined directly graphically.
However, clustering can be useful when examining relationships in such
situations [27]. Given the small number of objects to be clustered in the first phase of the analysis (the ten startup reasons), Ward’s hierarchical clustering procedure was used to obtain the dendogram shown in Figure 1.
There seem to be four distinct groups of startup reasons. The first were related to various economic and lifestyle aspects of running a small business (e.g., earning a living, independence), while the second were related to the creative aspects of small business. The third was a concern with being with the family and the fourth was a concern with using the small business as a form of employment.
work for self òûòòòø independence ò÷ ó be creative òòòòòüòòòòòòòòòòòòòø more money òø ó ó make a profit òôòòò÷ ùòòòø earn living ò÷ ó ó identified gap òûòòòòòòòòòòòòòòòòò÷ ùòòòòòòòòòòòòòòòòòòòòòòòòòø meet need ò÷ ó ó home with family òòòòòòòòòòòòòòòòòòò÷ find job òòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
ó
Figure 1: Dendrogram of startup reasons using Ward’s Method
In the second phase of the cluster analysis, respondents’ “scores” on the six retained dimensions were used to develop a “startup profile” for each small business. Following the procedure suggested by Green, Schaffer and Patterson [8] and Soutar and McNeil [26], respondents’ profiles (based on the six retained dimensions) were cluster analysed to find if there were distinct groups of firms with different startup patterns. The row results (respondents) were used in a further cluster analysis to see if there were groups of firms with similar export patterns. As there were a relatively large number of respondents (378), Howard and Harris’s K-means clustering procedure was used [13]. The number of cluster was varied from two to eight and, using the point biserial correlation [26], a six group solution was found to be the most appropriate, obtaining a point biserial correlation of 0.55. The percentages of respondents from each group who started their business for the various reasons are shown in Table 2
From Table 2 it is possible to develop an understanding of the mix of reasons why each of the six groups of small business operators began their businesses. Group 1 was the largest group (32% of the sample) and members had been motivated because they saw a gap in the market from which they could make a living. They were not greatly motivated by the other factors. Group 2 was also a large group (30% of the sample) that included people who were motivated by a perceived gap. However, this group wanted more than a mere living. They were also motivated to start their business because they believed that it would enable them to earn more money than they could earn elsewhere and that they would make a good profit by investing in such a business. They also were motivated by a desire to be independent. Group 3 members (10% of the sample) were almost totally motivated by a desire to be independent. Group 4 was the smallest group (3% of *the sample) and had very few
distinguishing motivations, although some (42%) had started their business because they were tired of working for some one else.
Group 5 (8% of the sample) were very similar to group 2 with one exception. While group 5 members were economically motivated, all were also motivated to start their business so that they could be at home with their family, a factor that differentiated them from all of the other groups. Members of Group 6 (17% of the sample) were motivated more by the chance to be creative than by any of the other factors, although they had also seen a market gap and an unfilled need than they believed would enable them to survive and, hopefully, prosper.
Table 2: Reasons for Startup – by Group Reason for startup
Group 1
Group 2
Group 3
Group 4
Group 5
Group 6
Greater independence and flexibility
13
67
85
33
77
34
To make a profit
32
81
8
25
68
14
To earn a living
49
51
10
25
65
22
To earn more money
13
95
28
25
77
46
Tired of working for some one else
23
46
18
42
61
38
More opportunity to be creative
6
12
8
33
39
75
Saw gap in the market
60
66
17
33
84
52
To meet a particular need in my area
1
6
5
17
32
66
To be home with the family
1
0
0
8
100
2
Could not find a job
0
0
51
0
9
0
The final stage of the analysis attempted to determine whether there were background differences between the six groups. Information on a number of often collected background variables (eg gender, education and family type) were included in the survey and examined in this phase of the analysis. The discriminant analysis revealed two significant functions that, using the I squared statistic suggested by Peterson and Mahajan [19], explained 26% of the variance between the groups.
The summary results are shown diagrammatically in Figure 2. The structural correlations, which are drawn as vectors in the diagram, show the relationship between the use of the various background variables and the estimated discriminant functions. The head of the vector shows the direction of the relationship, while the length shows the strength of the relationship. As can be seen from the diagram, groups 1 and 2 had very similar backgrounds and, indeed, the two groups were not statistically different from each other across the background variables collected in the present survey. Their families were more reliant on their small business for income and they were more likely to be involved in the business in a full time capacity. These seem to be sensible outcomes as both groups had been economically motivated to start their businesses.
Business major source of income * Group 2
Age
* Group 1
* Group 3
Involved full time Dependent Children
Female * Group 4
Figure 2: Background Differences Between Segments
Members of groups 3, 4 and 6 also had very similar and not statistically different backgrounds.
These
respondents were older, less likely to be involved in the business on a full time basis, less likely to obtain the majority of their income from their businesses and less likely to have dependant children. Again these outcomes seem sensible as group members had been motivated by lifestyle, as well as economic factors when they decided to stat their businesses. As can be seen from the diagram, group 5 members were very (statistically) different from all of the other groups. They were more likely to be females (39% compared to 18% for the overall sample) and were more likely to have dependant children (87% compared to 64% overall). Again these outcomes seem sensible as, while this group had been motivated by economic factors, they were unique in wanting a business that enabled them to stay at home with their family.
CONCLUSIONS
The present paper used a different analytical approach [correspondence analysis] to examine the reasons why people start small businesses. The results support much of the prior research that suggests small business owners started their businesses for a variety of reasons, which are usually both economic and lifestyle related. What was also clear was that there were subgroups of small business owners who started their businesses for quite different reasons [eg wanted to be able to stay home with the family, saw a market niche that would be extremely profitable of would provide a reasonable living].
Further, the pattern of reasons was related to people’s
backgrounds, especially their gender and family situation and the importance of the small business to the family’s income.
Given the variety of reasons for starting a small business, it seems likely that different small business owners will evaluate their success in quite different ways and, at least for some small business owners, in quite different ways to the ways large businesses assess their success. As such, further research is needed to examine the ways in which owners evaluate success and whether the various groups identified in the present research evaluate success in different ways. Such information, when combined with the present results, would provide useful information to small business researchers and to government policy makers attempting to develop useful and effective programs to assist both the start up and continuance of small businesses.
REFERENCES
1.
Birley, S. & Westhead, P. “A Taxonomy of Business Start-up Reasons and their Impact on Firm Growth and Size,” Journal of Business Venturing (1994) 9, 7-31.
2.
Borzi, A. The Gender Finance Gap, Working Paper, Sydney: Borzi, Smythe Pty. Ltd. (1994) May.
3.
Brush, C.G. “Research of Women Business Owners: Past Trends, a New Perspective, Future Directions,” Entrepreneurship Theory and Practice, (1992) 16(4), 5-30.
4.
Buttner, E. H. and D. P. Moore “Women's Organizational Exodus to Entrepreneurship: Self -Reported Motivations and Correlates with Success,” Journal of Small Business Management, (1997) 35(1), 3446.
5.
Chaganti, R. & Parasuraman, S. “More Alike than Different? A Comparative Study of Goals, Strategies, Management Practices and Performance of Small Businesses Owned by Men and Women”, Proceedings of the Small Business Institute Director’s Association, (1999) February.
6.
Envick, B. & Langford, M. “Behaviours of Entrepreneurs: A Gender Comparison,” Proceedings of Association for Small Business & Entrepreneurship, Annual Meeting, Dallas, Texas, (1998) March.
7.
Flinders University of South Australia Women in Small Business: A Review of Research, Report prepared for Department of Industry, Science and Technology, The University: School of Commerce, (1995) November.
8.
Green, P.E., Schaffer, S.M. & Patterson, K.M. “A reduced space approach to the clustering of categorical data in market segmentation,” Journal of the Market Research Society, (1988) 30(3), 267290.
9.
Greenacre, M.J. Theory and Applications of Correspondence Analysis, London: Academic Press, (1984).
10. ---- “SIMCA: A Program to Perform Simple Correspondence Analysis,” American Statistician, (1986) 51, 230-231. 11. Hisrich, R.D., Brush, C., Good, D., & De Souza, G. “Some Preliminary Findings on Performance in Entrepreneurial Ventures: Does gender matter?” Frontiers of Entrepreneurial Research, (1996) April, Babson College. 12. Hoffman, D.L. & Franke, G. “Correspondence Analysis: Graphical Representation of Categorical Data in Market Research,” Journal of Marketing Research, (1986) 23(3), 213-227. 13. Howard, H. & Harris, R.D. A Hierarchical Grouping Routine; IBM 360/65 FORTRAN IV program, Philadelphia: University of Pennsylvania Computing Centre, (1966). 14. Kuratko, D.F., Hornsby, J.S. & D. W. Naffziger, D.W. “An Examination of Owner's Goals in Sustaining Entrepreneurship,” Journal of Small Business Management, (1997) 35(1) 24-33. 15. LeCornu, M.R. McMahon, R.G.P. Forsaith, D.M. & Stanger, A.M.J. “The Small Enterprise Financial Objective Function,” Journal of Small Business Management, (1996) 34(3), 1-14. 16. Mallette, H & McGuinness “Gender Differences in the Entrepreneurial Start-Up Process,” Paper presented at the International Council for Small Business Conference, (1999). 17. Meredith, G.G. & Barrett, P.G. Self-Employed Entrepreneurs in Queensland, Lismore: Southern Cross University, (1994).
18. Naffziger, D.W., Hornsby, J.S. & Kuratko, D.F. “A Proposed Research Model of Entrepreneurial Motivation,” Entrepreneurship Theory and Practice, (1994) 18(3), 29-42. 19. Peterson, R.A. and Mahajan, V. “Practical Significance and Partitioning Variance in Discriminant Analysis,” Decision Sciences, (1976) 7(4), 649-65. 20. Shane, S., Kolvereid, L & Westhead, P. “An Exploratory Examination of the Reasons Leading to New Firm Formation Across Country and Gender,” Journal of Business Venturing, (1991) 6, 431-446. 21. Still, L.V. & Guerin, C.D. Self-Employed Women: The New Social Change, Report to the New South Wales Women’s Advisory Council, (1990) January. 22. Still, L.V., Guerin, C.D. & Chia, W. “Self-Employed Women: A Sydney Survey,” Women in Management Series, Working Paper No. 13, School of Business, University of Western Sydney, Nepean, (1990) April. 23. Still, L.V. & Chia, W. “Self-Employed Women: Four Years On,” Women and Leadership Series, Working Paper No. 1, Edith Cowan University, Deputy Vice-Chancellery, (1995) February. 24. Still, L.V. & Timms, W. Women and Small Business: Barriers to Growth, Report prepared for the Office of the Status of Women, Department of Prime Minister & Cabinet, Canberra, (1997) June. 25. Still, L.V. & Timms, W. “Small Business in Western Australia: A Comparative Gender Study,” Centre for Women & Business Discussion Paper Series, No. 4, Graduate School of Management, The University of Western Australia, (1999). 26. Soutar, G.N. & McNeil, M.M. “Information Search for a Professional Service,” Journal of Professional Services Marketing, (1995) 11(2), 45-50. 27. Soutar, G.N. & Wright, K. “Export Markets: a Correspondence Analysis,” in Cadeaux, J. & Uncles, M. (eds) Marketing in the Third Millennium, Proceedings of Australian and New Zealand Marketing Academy Conference, Sydney, Australia, (1999) December [CD-Rom].
