‘real’ differences between family and non- family …€¦ · address the weaknesses and...
TRANSCRIPT
1
‘REAL’ DIFFERENCES BETWEEN FAMILY AND NON-FAMILY SMEs: A COMPARATIVE STUDY OF AUSTRALIA
AND BELGIUM
Max Smith
Lecturer (International Business) Flinders Business School
Flinders University GPO Box 2100
Adelaide South Australia 5001
Tel: +61 8 8201 3897 Fax: +61 8 8210 2644
Email: [email protected]
Flinders Business School Research Paper Series
07-06
ISSN 1441-3906
KEYWORDS : Family business; industry differences; country
differences; Australia; agency theory
2
‘REAL’ DIFFERENCES BETWEEN FAMILY AND NON-FAMILY SMEs: A COMPARATIVE STUDY OF AUSTRALIA
AND BELGIUM
ABSTRACT This study attempts to further the development of family business theory by providing for a more detailed understanding of the differences between family and non-family firms’ profitability, growth, exporting and networking behaviour. Utilising data from 2190 Australian SMEs, the study compares the Australian experience of differences between family and non-family firms with those found among Belgium firms. The Australian results are consistent with the growth and some of the networking behaviour found among Belgium firms, but not with their profitability and exporting behaviour. The study’s findings support the contentions that the differences between family and non-family firms may be less than many earlier studies have indicated and that industry differences and cross-national differences in corporate governance environments may lead to a variance in the differences found between family and non-family firms. It also demonstrates that the underlying theoretical rationale for a number of predicted differences between family and non-family firms appears flawed. A significant number of recommendations for future research are presented.
KEYWORDS : Family business; industry differences; country differences; Australia;
agency theory
The key deficiency in the present scholarly understanding of family businesses is the
lack of a rigorous integrated theory of the family firm. One of the most important
issues that must be addressed in order to develop such a theory ‘is how and why this
form or organization behaves and performs in a distinguishably different way from a
nonfamily firm’ (Chua, Chrisman & Steier 2003: 334). As a consequence, the past
two decades have seen numerous studies carried out that attempt to identify and
classify all manner of differences between family and non-family firms. As is usual
with the introduction of a new field of academic enquiry, the focus of family business
research contributions has evolved from theory generation to hypothesis testing; and
from relatively simple testing procedures to more sophisticated methodologies that
address the weaknesses and limitations of earlier efforts.
In order to determine the universal differences between family and non-family firms
necessary for theory advancement, two important contextual issues have become
apparent from earlier studies. The first is concerned with the differing institutional
context firms operating in different countries encounter; while the second is
3
concerned with methodologies to control for firm specific contextual differences
within a particular nation. The first recognises that differences between family and
non-family businesses may vary according to the corporate environment they operate
in; while the second recognises that it is necessary to compare family and non-family
firms of a similar nature in order to identify true differences between the two groups.
Given that most family business research has been carried out in North America and
the United Kingdom (Gomez-Mejia, Nunez-Nickel & Gutierrez 2001), studies using
firm-data from other countries are valuable simply by virtue of their relative scarcity.
Of greater importance however, is their ability to confirm or deny the universality of
the differences found between family and non-family firms from these two regions.
For example, while severe agency conflicts were found between majority family-
owned and minority shareholders in East Asian family firms (Faccio, Lang & Young
2001) these problems were not apparent in Anderson and Reeb’s (2003b) study of US
family firms. Similarly, while Gedajlovic and Shapiro’s (1998) study confirmed the
agency theory proposition that low concentrations of ownership in a firm will lead to
reduced performance in US and UK firms, they also found that this was not the case
for firms in France, Germany and Canada. In both cases, national differences in the
respective corporate governance environments were considered the cause of these
variations because they altered the agency dynamics between managers and
shareholders in these nations. However, the results from these studies also indicate
that more cross-national comparative studies of family businesses are needed.
Many of the early family business studies tended to employ methodologies that
simply compared family and non-family businesses from the population available to
them without consideration of the differing demographics of the sample firms. It
appears likely this was a consequence of the extreme difficulty most researchers find
in obtaining reliable information on family firms (Schulze, Lubatkin & Dino 2003).
This, however, led to concerns about the methodology appropriate for family business
comparative research and calls for researchers to control for context when comparing
family and non-family firms. As Westhead (1997) and Westhead and Cowling (1998)
note, a failure to do so can lead to the identification of sample rather than real
differences between family and non-family firms. In response, many recent family
business studies have begun utilising multi-variate statistical techniques that control
4
for context. See, for example: Anderson and Reeb 2003a, 2003b & 2004; Gedajlovic
and Shapiro 1998; Gomez-Mejia et al 2001; Graves and Thomas 2004; Randoy and
Goel 2003; Schulze, Lubatkin, Dino and Buchholtz 2001; Schulze et al 2003.
In this regard, Jorissen, Laveren, Martens and Reheul’s (2005) study is valuable
because it addresses both the contextual issues mentioned above. The study’s sample
is Belgium based and therefore outside the US/UK region, and it also utilises
multivariate statistical analysis. In fact it goes beyond this by applying two sets of
statistical analysis to the dependent variables under examination using the same data
set. The first analysis ignored context while the second controlled for it (size, age,
industry & location). The results unequivocally demonstrate that a failure to account
for context leads to sample rather than real differences between family and non-
family firms and that this generally translates into results showing significantly more
differences between the two groups than is actually the case. Among other things, the
study found genuine differences between family and non-family firms’ profitability
and export intensity and no differences between their growth and networking
characteristics.
This paper attempts to further the country specific knowledge on differences between
family and non-family firms by comparing the results from Jorissen et al’s (2005)
Belgium based study with those generated from Australian based data after applying
multivariate statistical techniques that control for context. It also attempts to address
what appears to be a significant deficiency in the family business literature; namely, a
lack of cross national comparative research. The paper is structured as follows. First,
testable hypotheses are derived from past studies. Second, the data set and
methodology employed are outlined. Third, the results of the statistical analysis are
presented, and finally, the implications of the results are discussed.
PAST RESEARCH & HYPOTHESIS DEVELOPMENT
The main theoretical framework utilized to explain the differences between family
and non-family firms is agency theory. This theory has a relatively long history with
its genesis probably traceable as far back as Berle and Means (1932). The theory
contends that the interest of principals and agents do not coincide and that in the
absence of appropriate incentives and/or sufficient monitoring, agents will attempt to
5
maximize their own utility, often at the expense of the principals’ utility (Jensen &
Meckling 1976). For modern corporations, agency theory has been applied to the
relationship between the firm’s managers and its shareholders. The theory argues that
while owners (shareholders) want to maximize profits, their agents (managers) may
prefer to engage in self-interested, nonprofit-maximizing activities. As such, the
firm’s performance, to some extent, is dependent on the ability of owners to
effectively monitor and control managers (Gedajlovic & Shapiro 1998) and the cost
of carrying out these activities are known as agency costs.
Traditionally, agency models were concerned with the separation of ownership and
control in widely held firms (Gomez-Mejia et al 2001) where large information
asymmetries provided the greatest opportunity for managers to pursue their own self
interest. However, in the context of family businesses, where the degree of separation
between ownership and control is much less or non-existent, early agency theorists
concluded that agency problems and their associated costs were essentially non-
existent (Randoy & Goel 2003; Schulze, Lubatkin, Dino & Buchholtz 2001; Schulze
et al 2003). As Gomez-Mejia et al (2001: 83) note: ‘Jensen and Meckling (1976)
assumed that the blurring of the boundary between principal and agent in this type of
family contracting would make moral hazard largely inconsequential’.
However, in more recent times, a number of studies have emerged showing that
although ‘traditional’ agency costs appear to be reduced in family firms, there are
other family firm specific agency problems that arise to replace them. Among other
things, these agency threats can arise because the interests of the executive may not
necessarily coincide with those of the extended family; the lack of separation between
ownership and control leads to reduced formal safeguards; family involvement leads
to adverse selection of employees; there is a higher likelihood of ‘hold-up’ in family
firms (where one or more family members has a position of power over other
decision-makers and uses it to his/her own advantage); the firm experiences longer
lasting conflicts because protagonists are not willing to quit due to the higher exit
costs in family firms; family members have the ability to consume perks in order to
reduce residual claims of minority shareholders; there is high managerial
entrenchment in family firms; family managers are often sheltered from business risk;
family firm owners prefer lower business risk; altruism towards family members can
6
lead to inefficiencies; family members often have non-economic preferences and a
preference for cash flows over other options such as shareholder value or growth
(Anderson & Reeb 2003a; Gomez-Mejia et al 2001; Schulze et al 2001 & 2003).
As such, it would appear that agency problems, although perhaps of a different nature,
may still be significant in family firms and the costs associated with these may be
equal to or outweigh the benefits of concentrated ownership in reducing ‘traditional’
agency costs.
Profitability
Traditional agency theory predicts family firms will outperform non-family firms
because the separation of ownership and control allows non-family firm managers to
maximize their own utility function at the expense of firm profits (Demetz & Lehn
1985; Daily & Dollinger 1992; Randoy & Goel 2003). In addition, the personal and
emotional stake family members have in their firm makes them more committed to
business success (Davis 1983). However, as outlined above, recent studies have
indicated that family businesses are not without their own agency problems. Thus,
the determination of which type of firm is most affected by their respective agency
costs is a matter for empirical analysis. This task is made more onerous by the need
to consider the contextual issues mentioned earlier; namely, national differences in
corporate governance environments and demographic differences between firms in
their own nation.
Given these contextual issues, it is perhaps not surprising that prior studies in this area
have produced conflicting results (Randoy & Goel 2003). An examination of the
results from studies using methodologies that accounted for context among their
sample appear to show that national differences in the corporate governance
environment matter. For instance, Westhead and Cowling’s (1997) British based
study, after accounting for a large degree of context, found no difference between the
profitability levels of family and non-family businesses. In contrast, Anderson and
Reeb’s (2003a) and (2004) US based studies both found that, on average, family firms
perform better than non-family firms. In fact, their (2003a) study quantified the
difference showing that on average, US family firms have a 6.65 percent higher return
on assets (ROA) than non-family firms. Jorissen et al’s (2005) Belgium based study
7
found different results again, reporting that Belgium family firms have a lower net
ROA than non-family firms.
There are some methodological issues and inconsistencies that make comparisons of
these studies’ results difficult however. For instance, as already mentioned,
Gedajlovic and Shapiro’s (1998) five nation study found that diversified ownership
had a negative relationship with firm performance in the US and UK but not in
France, Germany or Canada. Given that family firms are typified by more
concentrated ownership, these results are consistent with Anderson and Reeb’s
(2003a & 2004) US based results but not with Westhead and Cowling’s (1997) UK
based study. However, this inconsistency may be a consequence of differences in the
samples and control variables utilized. For example, although Westhead and
Cowling’s (1997) study controlled for industry, firm age and location, it did not
control for firm size; and while Anderson and Reeb’s (2003a & 2004) studies
controlled for firm size, their sample was drawn from the S&P 500 firms. That is,
small and medium sized firms were excluded from the study. Similarly, the sample
used in Gedajlovic and Shapiro’s (1998) study was medium to large firms with assets
greater than US $50 million. In contrast to these earlier studies, Jorissen et al’s
(2005) sample included small to medium enterprises (SMEs) and also controlled for
firm size, making it much more comparable to the sample and methodology used in
the present study. As such, the following hypothesis is proposed.
Hypothesis 1. Australian family firms will have a significantly lower net ROA
than non-family firms.
Growth
According to early agency theorists, growth will often not be the primary objective of
family firms due to the overriding need to retain control of the firm for the family;
while for non-family firm managers, growth provides opportunities for higher
executive compensation and promotion (Daily & Dollinger 1992). More recent
studies have added to this expectation of lower growth from family firms by
highlighting how some aspects of observed family firm behaviour should affect
growth negatively. For instance, family firms’ preference for consumption in lieu of
investment is an oft cited reason to expect lower growth in family firms (Anderson &
8
Reeb 2003b; DeAngelo & DeAngelo 2000; Schulze et al 2003), as is the prevalence
of managerial entrenchment in family firms (Gomez-Mejia et al 2001) and their
preference for lower business risk (Anderson & Reeb 2003b; Casson 1999; Graves &
Thomas 2004).
Although Gallo’s (1995) findings supported these expectations, studies by Teal,
Upton and Seaman (2003) and Daily and Dollinger (1992) found no differences
between the growth of family and non-family firms in their samples. In addition, the
few empirical studies examining family business growth that control for context
within their sample have also found no differences in growth between the two groups.
Both Westhead & Cowling’s (1997) UK study and Jorissen et al’s (2005) Belgium
based study found no differences between the growth levels of their family and non-
family businesses. Given the weight of empirical results, these findings suggest that
the theoretical precepts underlining the expectation of lower growth for family firms
is in need of refinement and that the following hypothesis reflects our expectation;
particularly given the greater comparability between Jorissen et al (2005) and the
present study.
Hypothesis 2. There will be no significant difference between the growth levels
of Australian family and non-family firms.
Export Orientation
According to Graves and Thomas (2004: 8), ‘[f]amily business literature argues that
the complexities unique to family firms influence the attitude towards, and the extent
of, internationalisation’. This literature argues that the export orientation of family
firms is less than non-family firms because family firms are less growth orientated,
less involved in networks, more risk averse and prefer to live in close proximity to
their operations. Family firms are also likely to have cultures that are inward looking,
resistant to change and where decision-makers are constrained by the firm’s history
and tradition (Graves & Thomas 2004).
The empirical results from prior studies examining the differences between family
and non-family firms’ export orientation show little consensus. For instance,
Donckels and Frohlich’s (1991) study supported the literature with it’s finding that
9
family firms are less export orientated than non-family firms, while Gallo’s (1995)
study found no differences between the two groups in this regard. Similarly,
empirical studies investigating the export orientation of family firms that also control
for context show mixed results. Westhead and Cowling’s (1997) UK study, for
example, found no differences between the two groups while both Graves and
Thomas’ (2004) Australian based study of manufacturing SMEs and Jorrisen et al’s
(2005) Belgium based study found family firms export less than non-family firms.
Given that Westhead and Cowling’s (1997) study does not control for firm size,
probably an important consideration in this case, and the afore mentioned greater
comparability between Jorissen et al (2005) and the present study, the following
hypothesis is adopted.
Hypothesis 3. Australian family firms will have a significantly lower level of
exports (as a percentage of sales) than non-family firms.
Networking
Davis (1983) argues that the family firm’s relationship to the business environment is
affected by its inward orientation and its resistance toward information not matching
the family’s core beliefs. Both of these items would suggest that family firms
purposely engage in less networking activities than non-family firms. In addition,
according to Graves and Thomas (2004: 10-11), ‘Ward, 1997, and Okoroafo, 1999,
found that the majority of family firms were not aware of networks, such as
government programmes, that would assist them in internationalising’. That is, not
only are family firms less interested in networking, they are also less knowledgeable
about the networking opportunities available to them. This further reinforces the
argument that family firms are less involved in networking activities.
Donckels and Frohlich (1991) study, which found that family firms are less active in
networks, supported this contention, as did Graves and Thomas’ (2004) Australian
based study. However, in the later case, Graves and Thomas’ (2004) sample only
contained firms from the manufacturing industry and only examined ‘formal
networking with other firms’, thus it didn’t address a raft of other possible
information sources that could be included in a firm’s network. In contrast to these
studies, Jorrisen et al’s (2005) Belgium study found no significant differences
10
between the number of different types of contacts family firms use relative to non-
family firms. However, they also found that family firms are likely to have greater
frequency of contact with members of their network than non-family firms, which
tends to conflict with the afore mentioned studies’ theoretical contentions and
empirical results. Given that Jorrisen et al’s (2005) study examined more than one
industry and also provided for up to eight different types of contact, it once again
provides greater comparability with the present study. Hence, the following
hypotheses are proposed.
Hypothesis 4. There will be no significant difference between the number of
different types of contacts Australian family and non-family firms use in their
networking behaviour.
Hypothesis 5. Australian family firms will have significantly more frequent
contact with members of their network than non-family firms.
Table 1 provides a summary representation of the hypotheses discussed.
[insert Table 1 about here]
DATA SET & METHODOLOGY
The panel of firms utilized in this study consists of 2190 Australian SMEs (firms
employing up to 200 people) who had each provided four consecutive years of data
via Australia’s Business Longitudinal Survey (BLS). This survey was undertaken by
the Australian Bureau of Statistics (ABS) for the financial years 1994-95 to 1997-98
on behalf of the federal government. The integrity of the survey was enhanced by the
employment of various imputation techniques, including matching with other data
files available to the ABS, to deal with any missing data. In addition, because the
Australian Statistician could legally enforce the provision of appropriate responses to
questionnaires, response rates were very high by conventional research standards –
typically exceeding 90 per cent. The integrity of the panel was enhanced by ‘cleaning
up’ the data, that is, by omitting those cases containing logical inconsistencies. The
11
homogeneity of the panel was enhanced by only including those firms organized as
proprietary companies. Finally, the Mining industry and the Personal & Other
Services industry were not included in the panel due to an insufficient number of
cases.
Definition of Variables
Independent variables
For the purposes of this study a firm is classified as a family business if it meets three
criteria. The first is that put forward by Gasson, Crow, Errington, Hutson, Marsden
and Winter (1988) and Ram and Holiday (1993); namely, ‘whether members of an
“emotional kinship group” perceive their firm as being a family business’ (Westhead,
Cowling & Howorth 2001: 370). Thus, firms who answered in the affirmative to the
BLS question: ‘Do you consider this business to be a family business?’ were
considered to have satisfied this criterion.
The second criterion relates to another definition of family business used by other
researchers; namely, ‘whether a firm is managed by members drawn from a single
dominant family group’ (Westhead et al 2001: 370). This criterion was considered
met if a firm responded in the affirmative to the BLS question: ‘Do you consider this
business to be a family business because family members are working directors or
proprietors?’
The final criterion addresses another method used to distinguish family businesses
from other firms; namely, does the family in question hold more than 50 per cent of
the shares in the firm (Chromie, Stephenson & Monteith 1995; Westhead et al 2001).
This criterion was considered satisfied if a firm reported greater than 50 per cent of
it’s equity was held by family, either working or non-working, for at least one of the
years of the survey.
This study’s use of multiple criteria as the determinants of a family business
classification would appear to place these firms in the middle of Astrachan and
Shanker’s (2003) Family Business Universe and Westhead and Cowling’s (1998)
Scale of Family Firm Activity. Firms who didn’t meet these multiple criteria were
12
classified as non-family businesses. The resultant breakdown of the panel’s family
and non-family businesses across industry sectors is shown in Table 2.
[Insert Table 2 about here]
The independent variable for firm size is categorical and composed of Microfirms
(businesses that have 1 to 10 average total employees); Small firms (businesses that
have 11 to 49 average total employees); and Medium firms (businesses that have 50
to 200 average total employees). The independent variable for firm age is also
categorical and composed of firms whose age is either 2 to < 5 years old, 5 to < 10
years old, 10 to < 20 years old or 20 years old or greater.
Dependent variables
The first five dependent variables shown on table 5 (appendix) relate to profitability,
growth and export orientation and mimic those utilised by Jorissen et al (2005).
Profitability is captured by the four-year average net return on assets; growth is
captured by the three-year average yearly growth of employment, total assets and
gross profit; while export orientation is captured by the four-year average of the
sample firms’ export income as a percentage of sales. The results from each of these
five variables was divided into five ordinal categories to enable the use of ordinal
regression statistical analysis. It should be noted that Jorissen et al’s (2005) study
examined ‘value added’ in lieu of gross profit. Value added is not reported in
Australian financial statements and gross profit was considered a reasonable proxy in
this circumstance.
The remaining dependent variables relate to the firm’s networking activities, once
again in line with Jorissen et al’s (2005) study. However, in contrast with Jorissen et
al (2005), this study examines the frequency of contact with each of the ten types of
contacts available as this more detailed information is considered useful to the area
under investigation. In contrast to the first five dependent variables, data for the
networking variables only comes from the responses from firms in the 1995/96
financial year. This being the only year in the life of the BLS these questions were
asked. The sixth dependent variable, ‘Number of types of contacts’, was derived
from the total number of different types of contacts (out of the ten that follow) firms
13
had engaged with during that year, then once again dividing the results into five
ordinal categories to enable the use of the ordinal regression technique. The data for
the remaining ten dependent variables (Frequency of contact with…) was already
arranged into three ordinal categories within the BLS.
Statistical Method
The underlying continuous data within the BLS used to generate the variables
described above has irregular distributional properties (that is, it is non-normally
distributed). The transformation of metric variables to produce normal distributions is
avoided because of the difficulties of interpretation often created by such procedures.
Consequently, non-parametric/distribution free techniques of statistical analysis are
employed exclusively; thus ordinal regressions are used to determine statistically
significant differences between family and non-family businesses after accounting for
firm size, age and industry. The use of such methods is consistent with the vast bulk
of other studies that have utilised the BLS data (see for example, Graves and Thomas
2004; Johnsen and McMahon 2005a&b; Jones 2004a&b, 2005 & 2005/2006; Jones
and Xydias-Lobo 2004; McMahon 1999a&b, 2001a,b,c&d, 2003 & 2004; Smith
2003, 2005, 2006 & forthcoming) and the large sample size should account for any
perceived limitations in the ‘power’ of this technique when compared to higher order
parametric methods (Seigel & Castellan 1988).
From the previous section it can be seen that the independent variables in this study
are family/non-family business (FB/NFB), firm size (3 categories) and firm age (4
categories); while the dependent variables are Net ROA, three measures of growth,
exports as a percentage of sales, number of types of contacts and frequency of contact
with ten network members (see table 5 - appendix) The interest of this paper is in
determining whether being classified as a family business (relative to non-family
businesses) impacts on the outcomes of the sample firms’ profitability, growth,
exporting and networking activities. This is appropriate because the classification
(family or non-family firm) is a characteristic of the company, whereas the results
from these activities are variable outcomes. If family/non-family business is found to
be a ‘significant’ independent variable, this implies that this characteristic
significantly impacts on the outcome of the activity in question (i.e. the dependent
variable). In addition, because the ordinal regression reference category for this
14
variable is non-family businesses, any outcomes associated with a particular activity
can be attributed to family businesses.
The initial regression model employed by this study included ‘industry’ as an
independent variable in lieu of subdividing the sample into industry groups as shown
in table 5 (appendix). Given the results of Smith’s (forthcoming) study however, an
‘interaction’ variable (family business by industry type) was also incorporated into
this initial model. For each of the dependent variables, the output of this model
showed a high level of discord between the results of the numerous combinations of
industry type and family business, both in terms of statistical significance and the
direction of differences (positive or negative). Variation of this nature indicates it is
inappropriate to examine family businesses within the framework of all industries (as
an independent variable); instead it is necessary to utilise models that examine family
businesses on an industry basis. That is, the FB/NFB variable behaves so differently
in each industry it is necessary to examine each industry separately. Accordingly,
separate analysis was carried out for each industry, each of which focused on whether
family businesses differed significantly from non-family businesses within the
framework of multivariate regression model that controlled for size and age of the
firm.
Note that in the interest of conciseness, the results for the age and size independent
variables are not shown in table 5 (appendix). Although necessary for the
methodology, that is, to ensure that these contextual variables are controlled for, the
actual figures associated with them are irrelevant to the analysis carried out here.
Finally, in accordance with the use of SPSS statistical analysis software, the negative
log-log link function was utilised for the ordinal regressions after distributions of the
variables’ values indicated that lower categories were more probable (SPSS, 1999).
RESEARCH FINDINGS
Profit, Growth and Export Differences
Table 3 provides a simplified version of the profit, growth and export differences
between family and non-family businesses shown in the more detailed results of table
5 (appendix). As can be seen, family businesses are only likely to have a significantly
lower net ROA if they are from the Construction industry. If they are from the
15
Manufacturing or Wholesale Trade industries they are likely to have significantly
higher net ROA, and there is no significant difference between family and non-family
firms in this regard for the rest of the industries.
There are no significant differences between family and non-family firms’
employment growth across all industries and only family firms from the
accommodation industry show any significant difference in total asset or gross profit
growth. In both cases, family firms are likely to have significantly lower growth than
non-family firms.
The only difference between family and non-family firms’ export orientation occurs
in the manufacturing industry where family firms are likely to have a significantly
lower level of exports as a percentage of sales compared to non-family businesses.
[insert Table 3 about here ]
Differences in Networking
Table 4 provides a simplified version of the differences between family and non-
family businesses’ networking behaviour than that shown in the more detailed results
of table 5 (appendix). As can be seen, there are no significant differences across
industries between the two groups in relation to the number of different types of
contacts they utilized throughout the year.
Significant differences between family and non-family businesses’ frequency of
contact with various members of their business network are industry specific and with
the exclusion of family and friends, is generally the exception rather than the rule.
However, where there is a significant difference, in all but two cases family firms are
likely to have more frequent contact with the respective players.
[insert Table 4 about here]
DISCUSSION AND CONCLUSION
In relation to firm profit, this study’s results only support Hypothesis 1 (lower net
ROA) for firms in the Construction industry while the results for the Manufacturing
16
and Wholesale Trade industries are the opposite to those hypothesized. However, for
the majority of industries, there is no significant difference between family and non-
family firm profitability. From an agency theory perspective this suggests that for the
Construction industry, family business type agency costs outweigh the benefits of
concentrated family ownership while the opposite is true for firms from the
Manufacturing and Wholesale Trade industries who appear to adhere more closely to
the tenets of ‘traditional’ agency theory. Overall, these results are not consistent with
Jorrissen et al’s (2005) findings among Belgium firms and are more consistent with
Westhead and Cowlings’ (1997) UK results. This may indicate that Australia’s
corporate governance environment is closer to that of the UK than it is to Belgium
(something to be expected given the common legal heritage between Australia and the
UK). Similar studies of a comparable nature are needed in other countries to help
clarify this issue. For instance, it would be interesting to compare family and non-
family business profitability data from other Commonwealth countries with these
results. A similar result would support the common legal heritage contention and if
the industry specific differences hold true, may also indicate universal differences
apply and strongly signal the need to investigate what it is that makes family firms
behave differently in these industries.
The results relating to Hypothesis 2 (no differences in firm growth) are much more
consistent with only two significant differences out of a possible twenty-seven
outcomes from the three growth related variables. This is consistent with Daily and
Dollinger’s (1992) and Teal et al’s (2003) US results, Westhead and Cowling’s
(1997) UK results and Jorissen et al’s (2005) findings in Belgium. The absence of
growth differences between family and non-family firms therefore appears to be
widespread across industries and across the globe. This is puzzling given that the
theoretical rationale for growth differences between the two types of firms seems
compelling. If family firms do indeed prefer consumption to investment and capital
spending, have higher levels of managerial entrenchment leading to poor decision-
making, and engage in lower risk enterprises (for a more comprehensive list of
growth limiting factors associated with family firms see Anderson & Reeb 2004),
then it would seem safe to assume these firms would also be characterized by lower
growth. As this is not the case, the validity of these assertions seems questionable
17
and these issues need to be revisited by future studies that control for the contextual
variables already outlined. If these assertions are validated, then further research
needs to be undertaken that seeks to explain this logical inconsistency. The centrality
of this issue to the family business/non-family business dichotomy would also appear
to provide a strong imperative for further comparative studies, of a nature similar to
the present study, to be carried out in other nations.
Hypothesis 3 (lower level of export orientation) is only supported for the
Manufacturing industry with no differences or indeed no exporting (Accommodation,
Cafes & Restaurants) for the rest of the industries. As would be expected, the results
for the Manufacturing industry are consistent with Graves and Thomas’ (2004) study
that also examined firms from this industry. In this regard, the results demonstrate the
danger of extrapolation of single industry findings to firms in other industries given
that there is no differences between family and non-family firms for the rest of the
industries. Unfortunately, even though the outcome for the manufacturing industry is
consistent with theoretical expectations, some of the underlying rationale for this
expectation appears questionable. As shown in tables 3 and 4, in the manufacturing
industry there is no difference between family and non-family firms’ growth variables
or number of different types of contacts. In addition, table 4 also shows that
manufacturing family firms have more frequent contact with five of the ten members
of their network than non-family firms in the same industry and this is the largest
ratio of more frequent contact for all of the industries. This suggests that lower
growth aspirations and less involvement in networking activities, as mentioned in the
literature, is unlikely to be the cause of the result in this industry and future empirical
studies that control for context are needed to examine this anomaly. Viewed overall,
this result once again appears to have little consistency with Jorrissen et al’s (2005)
findings among Belgium firms and is more consistent with Westhead and Cowlings’
(1997) UK results.
While, Hypothesis 4 (no difference between the number of different types of contacts)
is confirmed universally across all industries, and is therefore consistent with Jorrisen
et al’s (2005) findings, the results for Hypothesis 5 (more frequent contact with
members of their network) is very much industry specific. Although it is important to
empirically confirm it, it is perhaps not surprising to find that contact with family and
18
friends is the only case where family firms are likely to have significantly more
frequent contact across the majority of industries. This majority of six out of nine is
improved to seven (Transport industry) if the results are examined at the 10 per cent
level of significance. The fact that in the vast majority of cases where there is a
significant difference, the direction of this difference is positive, tends to support
Hypothesis 5, although it could hardly be described as universal. For example, there
is no difference between family and non-family firms’ frequency of contact with
solicitors and consultants across industries, even after size and age of the firm are
accounted for. Similarly, there is no difference between family and non-family firms’
frequency of contact with all ten types of contacts for the Construction, Transport and
Finance industries. The manufacturing industry is the only industry with results that
could be interpreted as indicating family firms have more frequent contact with
members of their network, once again highlighting the danger of extrapolating
findings from one industry to others. Given there are few differences between family
and non-family firms’ networking activities and where there is, it is generally in a
positive rather than negative direction, then the inward orientation ascribed to family
businesses in the literature would appear questionable and studies designed to address
this inconsistency are needed.
From a methodological perspective, this study strongly supports the need to examine
differences between family and non-family firms on an industry basis as outlined in
Smith’s (forthcoming) study and further, to control for context within these industry
groupings. For the dependent variables used in this study, there is not one incidence
of a significant difference between family and non-family firms that spans all
industries. On the contrary, once size and age of the firm are controlled for, there are
a number of instances where no differences or very few differences between family
and non-family firms occur across industries. This supports Jorrisen et al’s (2005)
and Smith’s (forthcoming) contention that the number of differences between family
and non-family businesses, at least in terms of outcomes, are not a many as we have
been led to believe by past research.
From a theory building perspective the results discussed above provide a number of
instances where the theoretical rationale underlying predicted differences between
family and non-family businesses appears flawed. Therefore, new empirical studies
19
of the type mentioned above are urgently needed to address these issues and ensure
that the scholarly literature on family businesses is not being built on false
foundations. Of all the theoretical areas in need of attention, the most important
would appear to be the lack of growth differences between family and non-family
firms globally. Until this phenomenon is adequately explained, it would appear
unlikely that family business research will have a strong theoretical foundation to
build on.
On a more positive note, this study provides some support for the contention that
some differences between family and non-family firms may be affected by the
different corporate governance environment found in different countries. Empirical
studies designed to address this issue will be an important first step, however, if the
contention is confirmed, it is probable that subsequent studies gaining an
understanding of why these environments lead to different outcomes will be more
influential in the development of family business theory. Similarly, an understanding
of why outcomes vary between industries within a particular nation, may also lead to
theoretical advances. Such studies should allow for the determination of differences
in family firm behaviour that are persistent, even if they are found to only occur in
certain circumstances.
The present study has already provided a significant number of recommendations for
future research. However, in recognition of the dynamic nature of business
environments, contemporary studies in Australia are also needed in order to identify
behavioral changes and track subsequent trends. The discovery of changes to the
level and type of differences between Australian family and non-family firms during
the intervening period from the late 1990s, when the BLS data was collected, to the
present, would present an exciting opportunity. An understanding of why such
changes occurred would allow family business scholars to gain crucial insights into
the operation of family firms that should also lead to advances in theoretical
development.
In this regard, an apparent limitation of this paper is the age of the data utilized;
however, it is important to note that the BLS is a rare resource that may never be
duplicated. It provided this study with a sample size that is very large by family
20
business research standards together with data that has a high level of integrity. Such
a resource should be viewed very favourably given that 62 percent of family business
studies include no sample or a sample with less than 100 family firms, and 66 percent
of family business studies utilize convenience samples (Bird, Welsch, Astrachan &
Pistrui, 2002). Even the matched pairs methodology advocated by Westhead and
Cowling (1998), while increasing the efficacy of comparisons between family and
non-family businesses, does not allow for comparisons between different types of
family businesses unless a similarly large sample size is utilized. It would therefore
appear that the BLS’s ability to uncover greater depth of family business information
far outweigh its age limitation.
A more genuine limitation of this study is that it did not include large family and non-
family firms (those with greater than 200 employees) in its sample. It is possible,
indeed probable, that the behaviour of large family firms is different from smaller
family firms. If this is the case these differences may lead to variances from the
differences between family and non-family firms shown in the present study and
could well explain some of the conflicting results found in the literature. As such, the
present study’s findings can only be related to SMEs and further studies that cater for
the large firm context are needed.
In conclusion, this study has attempted to further the development of family business
theory by providing for a more detailed understanding of the differences between
family and non-family firms’ profitability, growth, exporting and networking
behaviour. The Australian results are consistent with the growth and some of the
networking behaviour found among Belgium firms, but not with their profitability
and exporting behaviour. The study’s findings support the contentions that the
differences between family and non-family firms may be less than many earlier
studies have indicated and that cross-national differences in corporate governance
environments may lead to a variance in the differences found between family and
non-family firms. Similarly, differences in the business environment between
industries appear to cause variances in the differences found between family and non-
family firms. The present study also demonstrates that the underlying theoretical
rationale for a number of predicted differences between family and non-family firms
is in need of revision.
21
Overall, this research is valuable to the study of family businesses because it:
• is one of the few studies to compare cross national differences between
family and non-family firms
• brings the Australian experience with profitability, growth, exporting and
networking differences between family and non-family firms to the attention
of scholars
• demonstrates where these differences vary with those found in Belgium (and
other countries)
• supports other findings that ‘real’ differences between family and non-family
firms are less than earlier studies would lead us to believe
• supports the need for methodologies that control for context, as advocated
and practiced by scholars recently and shows that national and industry
context need special consideration in this regard
• outlines a number of areas where the theoretical rationale for differences
between family and non-family businesses appears questionable
• provides a significant number of suggestions for the direction future research
should take in order to address these theoretical uncertainties and build a solid
theoretical foundation for the scholarly study of family businesses.
REFERENCES
Anderson RC and Reeb DM (2003a) Founding-family ownership and firm
performance: evidence from the S&P 500. The Journal of Finance 58(3): 1301-1328.
Anderson RC and Reeb DM (2003b) Founding-family ownership, corporate
diversification, and firm leverage. Journal of Law and Economics 46: 653-684.
Anderson RC and Reeb DM (2004) Board composition: balancing family influence in
S&P 500 firms. Administrative Science Quarterly 49: 209-237.
Astrachan J and Shanker M (2003) Family businesses’ contribution to the U.S.
economy: a closer look. Family Business Review 16(3): 211-219.
22
Berle AA and Means GC (1932) The Modern Corporation and Private Property.
MacMillan: New York.
Bird B, Welsch H, Astrachan J and Pistrui D (2002) Family business research: the
evolution of an academic field. Family Business Review 15(4): 337-350.
Casson, M (1999) The economics of the family firm. Scandinavian Economic History
Review 47:10-23.
Chua J, Chrisman J, and Steier L (2003) Extending the theoretical horizons of family
business research. Entrepreneurship Theory and Practice 27(4): 331-338.
Cromie S, Stephenson B, and Monteith D (1995) The management of family firms: an
empirical investigation. International Small Business Journal 13(4): 11-34.
Daily CM and Dollinger M J (1992) An empirical examination of ownership structure
in family and professionally manages firms. Family Business Review 5(2): 117-136.
Davis P (1983) Realizing the potential of the family business. Organizational
Dynamic 12: 47-56.
DeAngelo H and DeAngelo L (2000) Controlling stockholders and the disciplinary
role of corporate payout policy: a study of the Times Mirror Company. Journal of
Financial Economics 56:153-207.
Demetz H and Lehn K (1985) The structure of corporate ownership: causes and
consequences. Journal of Political Economy 93: 1155-1184.
Donckels R and Frohlich E (1991) Are family businesses really different? European
experiences from STRATOS. Family Business Review 4(2): 149-160.
Faccio M, Lang LH and Young L (2001) Dividends and expropriation. American
Economic Review 91: 54-78.
23
Gallo MA (1995) The role of family business and its distinctive characteristic
behavior in industrial activity. Family Business Review 8(2): 83-97.
Gasson R, Crow G, Errington A, Hutson J, Marsden T and Winter DM (1988) The
farm as a family business: a review. Journal of Agricultural Economics 39(1): 1-41.
Gedajlovic ER and Shapiro DM (1998) Management and ownership effects: evidence
from five countries. Strategic Management Journal 19: 533-553.
Gomez-Mejia, LR, Nunez-Nickel M and Gutierrez I (2001) The role of family ties in
agency contracts. Academy of Management Journal 44(1): 81-95.
Graves C and Thomas J (2004) Internationalisation of the family business: a
longitudinal perspective. International Journal of Globalisation and Small Business
1(1): 7-27.
Jensen MC and Meckling WH (1976) Theory of the firm: managerial behavior,
agency costs and ownership structure. Journal of Financial Economics 3:305-360.
Johnsen GJ and McMahon RGP (2005a) Owner-manager Gender, Financial
Performance and Business Growth amongst SMEs from Australia's Business
Longitudinal Survey. International Small Business Journal 23(2): 115-142.
Johnsen PC and McMahon RGP (2005b) Cross-industry differences in SME
financing behaviour: an Australian perspective. Journal of Small Business and
Enterprise Development 12(2): 160-177.
Jones JT (2004a) Training and Development, and Business Growth: a Study of
Australian Manufacturing SMEs. Asia Pacific Journal of Human Resources 42(1):
96-121.
Jones JT (2004b) HRM and Business Growth: an Exploratory Longitudinal Study of
Australian Manufacturing SMEs. Research Journal of the ooi Academy International
Congress 1(4): 75-87.
24
Jones JT (2005) The Determinants of Training in Australian Manufacturing SMEs.
Education + Training 47(8/9): 605-615.
Jones JT (2005/2006) New Management Practices and Training: a Longitudinal Study
of Australian Manufacturing SMEs. International Journal of Knowledge, Culture and
Change Management 5(4): 179-188.
Jones JT and Xydias-Lobo M (2004) Business Growth and Quality Management
Practices amongst SMEs: evidence from Australia's BLS. Research Journal of the ooi
Academy International Congress 1(4): 67-74.
Jorissen A, Laveren E, Martens R and Reheul A (2005) Real versus sample-based
differences in comparative Family Business research. Family Business Review 18(3):
229-246.
McMahon RGP (1999a) Financial reporting to financiers by Australian
Manufacturing SMEs. International Small Business Journal 18(1): 35-52.
McMahon RGP (1999b) Modelling the extent of financial reporting practices amongst
Australian manufacturing SMEs. Small Business Economics 13(2): 81-96.
McMahon RGP (2001a) Business growth and performance and the financial reporting
practices of Australian manufacturing SMEs. Journal of Small Business Management
39(2): 152-164.
McMahon RGP (2001b) Deriving an empirical development taxonomy for
manufacturing SMEs using data from Australia's business longitudinal survey. Small
Business Economics 17(3): 197-212.
McMahon RGP (2001c) Growth and financial profiles amongst manufacturing SMEs
from Australia's business longitudinal survey. Entrepreneurship Theory and Practice,
26(2): 51-61.
25
McMahon RGP (2001d) Growth and performance of manufacturing SMEs: the
influence of financial management characteristics. International Small Business
Journal 19(3): 10-28.
McMahon RGP (2003) An exploratory study of under- and over-investment amongst
manufacturing SMEs from Australia's business longitudinal survey. International
Small Business Journal 21(1): 29-53.
McMahon RGP (2004) Equity Agency Costs Amongst Manufacturing SMEs. Small
Business Economics 22(2): 121-140.
Okoroafo SC (1999) Internationalization of family businesses: evidence from
Northwest Ohio. Family Business Review 12(2): 147-158.
Ram M and Holiday R (1993) Relative merits: family culture and kinship in small
firms. Sociology 27(4): 629-648.
Randoy T and Goel S (2003) Ownership structure, founder leadership, and
performance in Norwegian SMEs: implications for financing entrepreneurial
opportunities. Journal of Business Venturing 18: 619-637.
Schulze WS, Lubatkin MH, Dino RN and Buchholtz AK (2001) Agency relationships
in family firms: theory and evidence. Organization Science 12(2): 99-116.
Schulze WS, Lubatkin MH, and Dino RN (2003) Exploring the agency consequences
of ownership dispersion among the directors of private family firms. Academy of
Management Journal 46(2): 179-194.
Siegel S and Castellan NJ Jr (1988) Nonparametric Statistics for the Behavioural
Sciences 2nd Ed. Singapore: McGraw-Hill.
26
Smith M (2003) From Entrepreneurial to Professional Management: a Transition
Model from Australian Manufacturing SMEs. Small Enterprise Research 11(2): 3-21.
Smith M (2005) Scanning the business environment and rates of growth among
Australian manufacturing SMEs. Small Enterprise Research 13(1): 22-36.
Smith M (2006) An empirical comparison of the managerial development of family
and non-family SMEs from Australia’s manufacturing sector. Journal of Enterprising
Culture 14(2): 127-143.
Smith M (forthcoming) Family businesses are not always different: An empirical
comparison of some managerial characteristics of family and non-family SMEs across
nine Australian industry sectors. International Journal of Entrepreneurial Behaviour
& Research.
SPSS Inc. (1999) SPSS Advanced Models 10.0. Chicago: SPSS Inc.
Teal EJ, Upton N and Seaman SL (2003) A comparative analysis of strategic
marketing practices of high-growth U.S. family and non-family firms. Journal of
Developmental Entrepreneurship 8(2): 177-195.
Ward JL (1997) Growing the family business: special challenges and best practices.
Family Business Review 10(4): 323-337.
Westhead P (1997) Ambitions 'external' environment and strategic factor differences
between family and non-family companies. Entrepreneurship & Regional
Development 9: 127-157.
27
Westhead P and Cowling M (1997) Performance contrasts between family and non-
family unquoted companies in the UK. International Journal of Entrepreneurial
Behaviour and Research 3(1): 30-52.
Westhead P and Cowling M (1998) Family firm research: the need for a
methodological rethink. Entrepreneurship Theory and Practice 23(1): 31-56.
Westhead P, Cowling M and Howorth C (2001) The development of family
companies: management and ownership imperatives. Family Business Review 14(4):
369-385.
28
Table 1 Summary of hypothesised relationships between Family Businesses (FB) and
Non-Family Businesses (NFB)
Table 2
Distribution of family and non-family businesses across industry sectors Family Business Non-Family Businesses Total Manufacturing 442 553 995 Construction 75 44 119 Wholesale Trade 171 251 422 Retail Trade 83 74 157 Accommodation, Cafes 11 36 47 & Restaurants Transport & Storage 32 50 82 Finance & Insurance 24 27 51 Property & Business 83 188 271 Services Cultural & Recreational 10 36 46 Services
TOTAL 931 1259 2190 (42.5%) (57.5%)
Profit & Growth
Export Networking
H1 H2 H3 H4 H5 FB More than NFB
Frequency of contact within network
FB Same as NFB
Growth
Number of types of contacts
FB Less than NFB
Return on Assets
Exports
29
Table 3 Simplified ordinal regression results for profit, growth and export differences
between Family and Non-Family businesses
Yes/No : Significant/Not Significant difference at 5% level +/- : Positive/Negative direction Control Variables : Firm Size and Firm Age
Industry M
anuf
Con
s
Who
lesa
le
Ret
ail
Acc
omn
Tra
nspo
rt
Fin
ance
Pro
pert
y
Cul
t & R
ec
Net ROA (%)
Yes +
Yes
-
Yes +
No
No
No
No
No
No
Employment growth (%)
No
No
No
No
No
No
No
No
No
Total asset growth (%)
No
No
No
No
Yes
-
No
No
No
No
Gross Profit (Value added) growth (%)
No
No
No
No
Yes
-
No
No
No
No
Exports (% of sales)
Yes
-
No
No
No
-
No
No
No
No
30
Table 4 Simplified ordinal regression results for differences between Family and Non-
Family businesses’ networking behaviour
Yes/No : Significant/Not Significant difference at 5% level +/- : Positive/Negative direction Control Variables : Firm Size and Firm Age
Industry
Manuf
Const
Wholesale
Retail
Accomodat
ion
Transport
Finance
Property
Cult & Rec
Number of types of contact
No No No No No No No No No
Contacts with Accountants
Yes +
No Yes +
Yes +
No No No No No
Contact with banks
Yes +
No No No No No No No Yes +
Contact with Solicitors
No No No No No No No No No
Contact with Consultants
No No No No No No No No No
Contact with Family and Friends
Yes +
No Yes +
Yes +
Yes +
No No Yes +
Yes +
Contact with Others in industry
No No No Yes +
No No No No No
Contact with Local Businesses
No No No Yes +
No No No No No
Contact with Industry Associations
Yes +
No No No No No No No No
Contact with Australian Taxation Office
No No No Yes -
No No No No No
Contact with Govt Small Business Offices
Yes +
No No Yes -
No No No No No
31
APPENDIX Table 5 (a)
Ordinal regression results showing where Family Businesses differ with Non-Family businesses across industries
(Reference category = Non-Family businesses)
Manufacturing Construction Wholesale Trade Dependent variables Wald Sig. Estimate Wald Sig. Estimate . Wald Sig. Estimate
1. Net ROA (%)
6.247 .012 .182 3.920 .048 -.435 7.057 .008 .306
2. Employment growth (%)
.881 .348 .076 .017 .896 .030 .290 .590 .070
3. Total asset growth (%)
.678 .410 .061 2.462 .117 .350 .345 .557 .070
4. Gross Profit (Value added) growth (%) 2.530 .112 .116 .154 .695 -.086 .065 .799 .030
5. Export (% of sales) 14.656 .000 -.395 .132 .716 .503 1.856 .173
-.240
6. Number of types of contacts .542
.462
.090
.054
.816
.087
.125
.724
.068
7. Freq. of contact with Accountants 39.931
.000
.628
.268
.604
.148
5.568
.018
.381
Frequency of contact with banks 21.437
.000
.568
.028
.866
.064
1.017
.313
.195
Frequency of contact with Solicitors .138
.711
.046
1.287
.257
-.438
.240
.624
-.095
Frequency of contact with Consultants 1.747
.186
.150
.026
.873
.071
1.468
.226
-.223
Freq. of contact with Family and Friends 42.028
.000
.695
2.499
.114
.507
12.855
.000
.649
Freq. of contact with Others in industry 1.797
.180
.117
.014
.907
.031
1.619
.203
.172
Freq. of contact with Local Businesses .022
.882
.018
.188
.665
.186
3.789
.052
.394
Freq. of contact with Industry Assoc’s 3.955
.047
.178
.228
.633
.155
.087
.768
.044
Frequency of contact with ATO .270
.604
-.051
.048
.827
.073
1.176
.278
-.161
Freq. Govt Small Business Offices 6.389
.011
.348
.347
.556
-.324
1.445
.229
.308
32
APPENDIX Table 5 (b)
Ordinal regression results showing where Family Businesses differ with Non-Family businesses across industries
(Reference category = Non-Family businesses)
Retail Accommodation Transport Dependent variables Wald Sig. Estimate Wald Sig. Estimate . Wald Sig. Estimate 1. Net ROA (%) .097
.755
.058
.130
.718
.164
.043
.835
.057
2. Employment growth (%)
2.469
.116
.355
2.518
.113
-1.119
2.009
.156
-.410
3. Total asset growth (%)
1.859
.173
.270
5.959
.015
-1.295
1.644
.200
.351
4. Gross Profit (Value added) growth (%)
.327
.568
-.107
11.777
.001
-1.896
.002
.962
.013
5. Exports (% of sales) .000
.994
-17.262
.
.
.
1.264
.261
-.897
6. Number of types of contacts
.693
.405
-.264
.065
.798
-.185
.007
.934
-.037
7. Freq. of contact with Accountants
7.861
.005
.694
2.318
.128
.880
.003
.960
-.018
Frequency of contacts with banks
.153
.696
.121
1.984
.159
1.077
1.644
.200
.585
Frequency of contact with Solicitors
2.161
.142
.468
.673
.412
-.659
1.452
.228
-.561
Frequency of contact with Consultants
.108
.743
.103
.771
.380
-1.009
1.093
.296
.521
Freq. of contact with Family and Friends
5.695
.017
.626
5.998
.014
1.906
3.058
.080
.702
Freq. of contact with Others in industry
17.382
.000
.912
.338
.561
.314
2.371
.124
.478
Freq. of contact with Local Businesses
6.796
.009
.810
.170
.024 .680
.266
.710
.399
.401
Freq. of contact with Industry Assoc’s
.221
.638
-.119
2.636
.104
-1.362
1.075
.300
.432
Frequency of contact with ATO
4.025
.045
-.506
.116
.733
.293
2.017
.156
-.658
Freq. Govt Small Business Offices
5.193
.023
-.977
1.991
.158
1.319
.010
.921
.075
33
APPENDIX Table 5 (c)
Ordinal regression results showing where Family Businesses differ with Non-Family businesses across industries
(Reference category = Non-Family businesses)
Finance Property Cult & Rec. Services Dependent variables Wald Sig. Estimate Wald Sig. Estimate . Wald Sig. Estimate 1. Net ROA (%) .577
.448
.279
.858
.354
-.145
.478
.489
-.295
2. Employment growth (%)
.322
.571
-.214
.001
.981
.004
.000
.992
-.004
3. Total asset growth (%)
.025
.874
.057
.963
.326
.150
2.459
.117
.674
4. Gross Profit (Value added) growth (%)
.011
.916
-.040
1.924
.165
.215
1.761
.184
.545
5. Exports (% of sales) .000
.998
.003
.076
.783
-.101
.001
.975
-.038
6. Number of types of contacts
.041
.840
.119
.048
.827
.057
.436
.509
.445
7. Freq.of contact with Accountants
3.819 .
.051
.878
.040
.841
-.038
2.545
.111
1.043
Frequency of contacts with Banks
.149
.699
-.245
.179
.672
.111
5.863
.015
1.988
Frequency of contact with Solictors
.142
.707
-.229
.751
.386
-.227
1.487
.223
1.013
Frequency of contact with Consultants
.473
.492
-.478
.367
.545
.179
1.598
.206
-1.086
Freq. of contact with Family and Friends
1.816
.178
-.812
4.162
.041
.441
9.429
.002
1.924
Freq. of contact with Others in industry
.017
.897
-.048
.001
.976
.005
.418
.518
-.291
Freq. of contact with Local Businesses
2.664
.103
-1.155
75
.600
.161
.499
.480
-.559
Freq. of contact with Industry Assoc’s
.278
.598
.309
.027
.869
.043
1.586
.208
-.954
Frequency of contact with ATO
.147
.701
.232
.005
.943
.016
2.737
.098
-1.388
Freq. Govt Small Business Offices
.507
.476
-.877
.341
.559
-.243
.000
.996
-16.547