A comparative study of cooperative relationships in family and non-family businesses

The aim of the present article is to find out differences and similarities in inter-firm cooperative relationships among family and non-family businesses, in order to contribute to an integrated theory on strategic alliances and family businesses within the small and medium-sized enterprises (SME) context. We found important divergences between both types of businesses, namely in the perceived difficulties and in performance evaluation. The results underpin the importance of trust when family businesses undertake cooperative relationships. An argument for cooperation between family businesses is made. We also present new research issues in family business and strategic alliance research.


INTRODUCTION
Faced with market globalisation, competitive pressures and technological changes, firms need to be flexible and are challenged to include these aspects in their strategy, in order to achieve the necessary competitiveness.Hereby, as referred to by Parkhe (1991), Kanter (1994) and Guinn and Young (1996), competitiveness no longer depends only on firms" internal capacity, but also on the relationships they establish with other firms.Resource constraints often make it impossible to bear the high costs of research and development as well as foreign market penetration and expansion, suggesting that cooperative relationships are important (Crick and Jones, 2000).
As Contractor and Lorange (1988) argued, a successful strategy often requires adoption of cooperative alliances that lead firms to obtain a competitive advantage.Alliances enable firms to complement their core competencies and to obtain additional resources; they have grown in popularity over the last years (Dyer et al., *Corresponding author.E-mail: mfranco@ubi.pt.Tel: +351 275 319 6636.Fax: +351 275 319 601. 2001; Augustine and Cooper, 2009).However, such arrangements are difficult to manage (Fryxell et al., 2002) and have high rates of failure (Kogut, 1989;Geringer and Hebert, 1991;Park and Ungson, 1997;Yan and Zeng, 1999).As referred to by Harrigan (1988), the dissolution rate is about fifty percent, and Kale et al. (2002) observed a similar rate of failure.
Firms form alliances in order to address promotional, operational, relationship, or strategic issues (Augustine and Cooper, 2009).In this paper, we define cooperative relationships as formal or informal alliances among firms with a strategic purpose.According to Teng (2007), strategic alliances can be of many types such as joint production, equity alliances, joint ventures, research and development (R and D) alliances and learning alliances, and these forms of cooperative relationships are not exclusive.
In spite of the importance of cooperative relationships for improving business competitiveness, research combining strategic alliances with family businesses is scarce.As remarked by Dyer (2003), a small number of studies in organisational research have included the family as a variable.On the other hand, in family business literature, only a few papers have dealt with exploring inter-firm cooperations in the context of networking family firms, for example Niemelä (2004) and Roessl (2005).
However, to our knowledge, no explicit research has yet been carried out to examine the specific characteristics of inter-firm cooperations comparing family and non-family businesses.This is particularly true for small and medium-sized enterprises (SME).Our research question, therefore, is: What are the similarities and differences in performing cooperative relationships between small and medium-sized family and non-family businesses?In an interview, Professor Danny Miller has claimed to compare family and non-family firms of a given type and to study family firms in the light of managerial theories for detecting their advantages and challenges (Moores, 2009).Taking these recommendations further, with our research we contribute to the construction of an integrated theory on strategic alliances and family businesses within the SME context.
The remainder of this paper is organised as follows.Subsequently, this study gives a literature overview of the research hypotheses and contributes to theory building.Then, it describes data collection, measures and statistical methods.A descriptive as well as explorative data analysis was then undertaken, after which robustness tests were carried out and the findings were discussed.Finally, the study was concluded and the implications were highlighted.

RESEARCH HYPOTHESES AND THEORY BUILDING
In the ensuing theoretical reflections, our intention is to shed light on the specific characteristics of family businesses compared to their non-family counterparts, in order to detect their impact on cooperative relationships.In particular, we first address the issue of whether family firms are likely to form cooperative relationships and why they might have difficulties compared to other types of businesses.Afterwards, we look for differences in objectives for cooperating and in performance evaluation.Based on these premises, we develop three research hypotheses and our research model.
Literature indicates that family businesses have a series of specific characteristics offering opportunities, but also difficulties for their engagement in cooperative relationships.Dyer (1986) speaks of a "family factor" that drives behaviour in family businesses.Mishra and McConaughy (1999) found that family goals and needs are often the deciding factors in strategic management decisions.Zahra et al. (2004) found that organisational cultures, i.e. coherent pattern of beliefs and values, have a greater strategic significance for family businesses than for non-family firms.Recently, Härtel et al. (2009) has reiterated the significance of noneconomic variables within the family business environment.
In several studies, trust-based organisational culture was reinforced as an outstanding quality and competitive advantage in family businesses (Lubatkin et al., 2005;Zellweger, 2007;Sundaramurthy, 2008).This goes along with parental altruism, which also permeates the dynamics of family businesses (Schulze et al., 2001;Lubatkin et al., 2005;Lee, 2006).Family members are potentially closer to the values and practices of the family firm than would happen in other types of business.Therefore, governance is different from non-family firms and altruism influences considerably the ability of the firm"s owner-manager to exercise self-control (Lubatkin et al., 2005).It is also true that strong family ties may lead to unbridled nepotism or serious interpersonal conflicts (Dyer, 1986).
At first glance, these features seem to meet the success factors for entering into cooperative relationships.As noted by Badaracco (1992), distrust among partners is an important variable when firms adopt cooperation processes.Since partners may behave opportunistically in alliances, trust or contractual safeguards between partners (Luo, 2002;Lui and Ngo, 2004;Augustine and Cooper, 2009) and political behaviour (Tushman, 1977;Papadakis et al., 1998) are essential requirements.As the difficulties in cooperative relationships are usually more of cultural than economic origins (Chau, 1991), trust and loyalty represent, therefore, two advantages of family businesses in inter-firm cooperations (Swinth and Vinton, 1993).
Furthermore, Niemelä (2004) suggests that knowledge, skills, motivation and volition, along with personal and institutional power; affect the ability for inter-firm cooperations.Therefore, as family businesses are run by owner-managers who combine of both characteristics, they seem to be predestined for cooperative relationships.
At the same time, the particular characteristics of family businesses may also hinder the formation of alliances.Cooperative relationships are affected by the cultural distance between partners.In fact, it is in a cultural environment that social interactions take place and no business could operate without these relationships.As noted by Tushman (1977), differentiation increases the difficulty to communicate and is a potential for conflict.
According to Posso (2003), family businesses are unique and closed systems with specific norms, values and languages, which create differentiation and may be contrary to good communication and interrelationships, especially when dealing with non-family businesses.
Moreover, comparing family firms to non-family firms, Astrachan and Kolenko (1994) note the limited organisational competencies in family businesses, especially in implementing proactive management practices such as strategic planning.Hienerth and Kessler (2006) state that their owner-managers are less prepared to assume internationalisation or cooperative strategies.Lee (2006) argues that family businesses tend to limit top management positions to family members rather than hire more qualified professionals from outside.This closeness might be contrary to cooperative relationships.This comes along with less structured (Kotey, 2005) and conservative management styles (McConaughy et al., 2001), which makes them react slowly to environmental changes (Daily and Dollinger, 1993).Because of family businesses" specific characteristics, Roessl (2005) even contends that they are less willing to enter into cooperative relationships.These reflections, despite some converse arguments, lead to our first hypothesis: H 1 : Due to the complexity of cooperative relationships, family businesses have more difficulties than non-family businesses in cooperating.
When trying to understand the reasons for cooperative relationships, generally speaking, there is a consensus about the objectives behind firms adopting such relationships that is, accessing innovative technology, entering into new markets, reducing production costs, increasing profits, etc. Hereby, resource-based theory (Pfeffer and Salancik, 1978;Wernerfelt, 1984) seems to be an appropriate theoretical perspective, as alliances are developed to manage scarcity of resources (Das and Teng, 2000).In the considerations stated in the foregoing, we hypothesised that family businesses have special needs and hindrances in inter-firm cooperations.This is particularly true for their imperfect organisational and managerial capacities.Cooperative relationships could be a suitable strategy to overcome these limitations, above all when firms compete in international markets.
For family businesses in particular, there are also other possible objectives for engaging in cooperative relationships.Family businesses usually have long-term visions as opposed to short-term profit goals (Daily and Dollinger, 1993;Harris et al., 1994;Zahra et al., 2004;Zellweger, 2007), related to continuity of the firm and valuing stability against risk (Donckels and Fröhlich, 1991;Daily and Dollinger, 1993).
According to Reid and Adams (2001), their preliminary objective is preserving ownership.The strategic horizon of family business-owners habitually spans generations (Ward, 1988;Sharma and Irving, 2005).Taking the argument further, Lee (2006) suggests that ownermanagers view the firm as an asset to pass on to future generations.In his study, Posso (2003) revealed that family firms use cooperative relationships to provide family members with better salaries or to decrease the need for investment.Whatever the specific objectives in family businesses for building alliances are, most of them appear to be family oriented.Based on these arguments, our second hypothesis is: H 2 : Family businesses have specific characteristics, so when involved in cooperative relationships, they pursue With regard to the evaluation of alliance performance in general, there is a variety of potential indicators, such as above all profitability, growth, costs or survival of the alliance, but also satisfaction with the cooperation and other "soft" measures.However, in family businesses, the definition and measurement of success is sometimes different from that of non-family businesses (Habbershon and Pistrui, 2002;Astrachan, 2003;Chrisman et al., 2003), which could also have implications on the evaluation of cooperative relationships.As already described in the foregoing paragraph, the former prefer strategies related to long term profits and are more riskaverse than non-family businesses, as business failure could have more serious consequences for them than for managers with little or no ownership interest (Daily and Dollinger, 1993).Growth, in family firms, occurs at a pace sufficient to meet the needs of family members in the firm (Daily and Dollinger, 1993), so that, in the main, family business owners-managers are less likely to pursue growth than non-family firms (Donckels and Fröhlich, 1991;Kotey, 2005;Hienerth and Kessler, 2006).If growth is intended, it is pursued with great caution (Upton et al., 2001), and ambitious goals for growth, including internationalisation, are likely to be pushed by second-generation proprietors (Timmons, 1994).Instead, it has been demonstrated that family businesses strive rather more for benefits unrelated to financial and competitive performance, that is, a strong preference towards non-economic outcomes (Dunn, 1995;Chua et al., 2003;Chrisman et al., 2004;Craig and Moores, 2005;Gómez-Mejía et al., 2007).For example, in family businesses, ownership transition and efficiency are often used to capture successful performance (Sharma et al., 1997;Habbershon and Pistrui, 2002).Hoffman et al. (2006) presented a new concept to explain the family influence on business performance: family relationships.In the same vein (Dyer, 2006), comparing the performance of family firms with firms having no family ties, proposes considering the "family effect" on organisational performance.These relations or effects shape the administration and firm"s structure, not least as ownership and direction are in the same group.The unique attributes of family firms; trust, loyalty and altruism, can therefore lead to a different evaluation of performance.Therefore, our third hypothesis would be: Summarizing, Table 1 depicts the research model on family businesses" idiosyncratic characteristics that might influence their inter-firm cooperative relationships.(2) a updated database that was used in a previous empirical study (Franco, 2001); and (3) identification of some cases in the EXAME review (Portuguese business Journal).From these sources, we extracted a convenient (Patton, 1990) sample of 627 Portuguese SME that had participated in cooperative relationshipsregardless of being family businesses or not.Inter-firm cooperation relationships were defined as a formal or informal relationship between two or more firms to pursue a set of common interests through the sharing of resources: joint ventures, strategic alliances, and other interorganisational forms.
Then we administered a questionnaire to the person in charge of the firms selected in the period from January to April 2006.After one reminder, 82 usable questionnaires were returned, representing a 13.1% response rate.To identify family businesses, there is a variety of definitions in the literature.They mostly have in common the use of property and management dimension.In particular, researchers have delimited a family business as one that is partly owned by one or more family members controlling at least 20% of the votes (La Porta et al., 1999;Miller and Le Breton-Miller, 2006).
In our study, we considered firms where at least 20% of the capital was with the founding family and at least one person in this family assumes top level management.In that way the family has perfect control of the firm.Applying these criteria, the sample resulted in 35 family businesses and 47 non-family businesses.

Measures
Firms" characterisation was measured in terms of activity sector, year of foundation, number of employees and level of internationalisation (cf.for the latter Calof and Beamish, 1995).The questionnaire also included questions about the three dimensions considered to evaluate cooperative relationships: (1) difficulties, (2) objectives and (3) performance.The firms were asked to value the items on a five-point Likert scale (that is, from 1 = "of no importance" to 5 = "of major importance").
An ordinal classification of perception is a more realistic task for respondents than the use of interval or ratio measures.It was also expected that managers would have only a limited amount of time to devote to the questionnaire, hence an easily understood Likert scale appeared to be more feasible than a potentially more precise but more complex scaling method.A five-point scale was adopted because it was felt that more numerous response categories would exceed the respondent"s ability to discriminate, with the likelihood that "noise" rather than more precise data would result.

Statistical analysis
A first approximation to the behavioural understanding of the firm sample was made from a frequency analysis of the variables included in the questionnaire.In this analysis, we made a Fisher"s exact test.This statistical test is computed when the sample size is small.Secondly, results were obtained from a factor analysis of variables indicating the main objectives, difficulties and performance indicators perceived by the firms regarding the interfirm cooperation processes.
To check acceptability of the technique, the Kaiser-Meyer-Olkin sample suitability measure, the Bartlett Sphericity Test and the Cronbach"s Alpha were taken into consideration to measure the level of consistency between variables.Afterwards, we made a ttest to analyze the differences in cooperative relationships concerning objectives, difficulties and performance indicators between the two groups (two independent samples); family and non-family businesses.

Sample characteristics
Table 2 reveals that both family businesses and nonfamily businesses were found in all industry sectors.However, besides machines and equipment, family businesses are more representative in the traditional areas textile and shoes and furniture and paper pulp.
Non-family businesses are more frequent in capitalintensive industries such as machines and equipment as well as chemical and plastic products, but were also often found in textile and shoe industry.However, these differences are not statistically significant.
Statistically significant differences between the two sub samples were found in other variables.The non-family businesses are olderabout one half (45.5%) were created before 1970, whereas the majority of family businesses (80%) were created after that year.Also, the family businesses in our sample are smaller: more than one half (54.3%) have fewer than 50 employees, while the largest fraction (70.2%) of non-family businesses has more than 50 employees.
Concerning sales for the international market, the data shows that the differences between the two groups are less important.The firms in the two sub samples have exporting practices at diverse levels, and the differences are not statistically significant.Subsequently, we will proceed with explorative data analysis and respective discussion.To do so, we analyzed inter-firm cooperations based on three dimensions: (1) difficulties, (2) objectives and (3) performance.

Difficulties in cooperation
In order to determine firms" main difficulties in participating in cooperative relationships, they were asked to value a set of variables (difficulties).The difficulties with cooperative relationships based on the mean measure are shown in Table 3.
In general, the firms did not give too much importance to the items in this dimension.The difficulties have means between 2.97 and 2.29 on a five-point scale except in the variable "time spent on the negotiation process", with a mean of 3.16.This might be explained by the fact that the high rates of failure pointed out by research (Harrigan, 1988;Kogut, 1989;Geringer and Hebert, 1991;Park and Ungson, 1997;Yan and Zeng, 1999;Kale et al., 2002) are the results of real problems.
From factor analysis, four factors were produced making reasonable conceptual sense and covering a total of 59.83% of the observed variance, as shown in Table 3.For each factor obtained from factor analysis, a reliability analysis was performed (Cronbach"s Alpha), in order to measure the internal consistency of each scale as a whole.Such an index, which varies between 0 and 1, measures the consistency among variables.According to Hair et al. (1998), the closer the index is to 1, the better the scale"s internal consistency.Here, the values of Cronbach"s Alpha show good stability among variables within factors.The remainder of this section discusses the interpretation of each of these factors.

Dependency and operational difficulties (factor 1)
Cooperative relationships help firms to access resources.According to the resource dependence perspective (Pfeffer and Salancik, 1978;Wernerfelt, 1984), inter-firm relations are developed to better manage a firm"s Notes: The mean is the average on a scale of 1 (="of no importance") to 5 (="of major importance").KMO Measure of sampling Adequacy = 0.733; Bartlett Test of Sphericity: 388.191; df = 105; p < 0.000.Fi (i=1…4) -Factors; Principal components factor analysis with varimax rotation.
dependence on resources (Das and Teng, 2000).However, the cooperation process requires some dependence and necessity of operational support.This factor includes precisely difficulties associated with the small size of the firm, fear of losing independence or autonomy, lack of financial capacity in the firm and legal and political aspects.

Opportunistic behaviour and fear (factor 2)
This factor aggregates some variables linked with the behaviour of alliance partners: Opportunism and a partner"s tendency to become a competitor.As exposed by Badaracco (1992), the question of opportunism and own interest is of particular importance when firms intend to create new knowledge and capacities.In this sense, there is a high potential for conflict and risk that the cooperation fails.According to Limerick and Cunnington (1993), inter-firm cooperations can also create a competitive climate, even to the extent that cooperation might "produce" new competitors out of former partners.Moreover, lack of confidence and reciprocity between partners can cause additional transaction costs.In short, the problems that accompany this factor appear to be related to the underlying and actual intentions of entering into cooperations.
Negotiation and coordination (factor 3) Slowinski et al. (1996) believe that many of the difficulties that occur during cooperation strategy implementation could have been identified during negotiations.This factor includes items related to "time spent on the negotiation processes, "different expectations" and "communication problems".As noted by Vyas et al. (1995), the initial agreement is frequently based on the mutual benefit that each firm is likely to obtain from the union of effort.As time progresses, one party may find that it no longer needs the skills or knowledge the other firm brought to the union because such knowledge and skills are duplicable, given the learning experience.

Distrust and communication (factor 4)
Building inter-partner trust is an important aspect of managing the relationship and deserves particular attention.If there is no mutual trust among partners, uncertainty about opportunistic behaviour by the other partner can hinder the beginning of negotiations or disrupt the agreement.Swinth and Vinton (1993) pointed out trust and loyalty as advantages of family businesses in inter-firm cooperations, but Reid and Adams (2001) point out that the fear of losing ownership and control of the firm represents an important constraint for cooperation in family businesses.Especially in the case of cooperative relationships, the influence of firms" culture is often crucial.When cultures cross, a cultural shock can occur (Meschi and Roger, 1994).The greater the cultural distance between partners in cooperation processes, the more evident the organisational effects of such a shock become.
The results of testing hypothesis 1 (family businesses have more difficulties than non-family businesses) are shown in Table 4.
Our findings indicate statistically robust differences between family and non-family businesses in all factors.In particular, family business show more difficulties in factor 1 (Dependence and operational difficulties), factor 2 (Opportunistic behaviour and fear) as well as in factor 4 (Distrust and communication), compared to non-family businesses.Contrariwise, non-family businesses seem to have relatively more difficulties with regard to factor 3 (Negotiation and coordination).With these results, hypothesis 1 is statistically supported.
A more detailed look reveals that, within factor 1, "small size of the firm", "lack of financial capacity in the firm", "geographical distance" and "high cost in the process" are the outstanding problems for the family businesses in our sample.This is in line with Astrachan and Kolenko (1994) and Hienerth and Kessler (2006), who note the limited organisational and strategic competencies of family businesses.In addition, "fear of the partner being a competitor" and "opportunistic behaviour by partners", main components of factor 2, are significantly more important for family businesses.Regarding factor 4, all the difficulties, that is, "internal constraints in the firm", "lack of trust between partners" and "cultural differences" were more frequently cited by family businesses.This underpins the importance of trust (Lubatkin et al., 2005;Zellweger, 2007;Sundaramurthy, 2008) when family businesses undertake cooperative relationships.

Objectives of cooperation
The main objectives for participating in cooperative relationships are shown in Table 5.Here, firms were asked to value a set of variables that were later submitted to factor analysis.For the full set of agreements, the mean is higher for the following objectives: "entering new markets", "to achieve competitive advantages" and "to consolidate market position".Other relatively highly Franco and Haase 10879 ranked objectives are "to improve the level of innovation" and "to create and explore synergies".It is clear from Table 5 that the highest ranked objectives are concerned with competitive positions in either new or existing markets, as well as in innovative activities.
As noted before, the 19 objectives represent a number of overlapping perspectives.From the factorial analysis, six factors were produced that make good conceptual sense.They also account for a total of 67.05% of the observed variance, as shown by Table 5. Observation of the values of Cronbach"s Alpha indicates that all the indexes are higher than 0.64, but not very close to 1.Each of these factors is interpreted in the ensuing considerations.

Innovation and organisational learning (factor 1)
This factor includes objectives associated with development of resources and competencies in the areas of technology and innovation, with benefits in terms of quality.According to Teng (2007), inter-firm alliances can greatly assist innovation efforts."To achieve some experience" was another reason for collaboration.
Experience is the main source of inter-organisational learning.Alliances are formed, above all, by partner organisations to learn from each other"s knowledge base (Khanna et al., 1998).They serve as the basis through which firms intensively interact and gradually absorb knowledge from their partners (Doz, 1996).This factor also aggregates the variable "to foment learning processes" as an important reason related to alliance formation.In fact, Hamel and Doz (1989) propose that alliances should be seen as learning opportunities, and several later studies have confirmed this point of view (Gulati, 1995;Eisenhardt and Schoonhoven, 1996).

Competitiveness and cost sharing (factor 2)
This factor includes variables related to the search for competitive positioning by the firm as well as risk and cost sharing.Concerning this last aspect, Contractor and Lorange (1988) showed that cooperative relationships can reduce a partner"s risk, i.e. the cost of the partnership is less than the cost of investment undertaken by each firm alone.A cooperative relationship can lower the total investment cost of a particular project or the assets at risk, by combining expertise and slack facilities in the parent firms (Glaister and Buckley, 1996).

Development and market power (factor 3)
This includes two main objectives with respect to the variables: "entering new markets" and "to facilitate international expansion".International collaboration can  Notes: The mean is the average on a scale of 1 (="of no importance") to 5 (="of major importance").KMO Measure of sampling Adequacy = 0.734, Bartlett Test of Sphericity: 534.550; df = 171; p < 0.000.Fi (i=1…6) -Factors; Principal components factor analysis with varimax rotation.
be the easiest way to penetrate foreign markets.For firms without international experience, an initial overseas expansion is often likely to be a cooperative relationship.Contractor and Lorange (1988) argue that, in general, it is an expensive, difficult and time-consuming business to establish a global organisation with a significant international competitive presence.In fact, the move to new foreign markets and the development of either a multidomestic or global strategy can be facilitated by cooperative relationships even for firms with considerable overseas experience.

Resource dependence (factor 4)
When firms collaborate, they can achieve resources not otherwise available (Teng, 2007).Hereby, Das and Teng (2000) developed a resource-based theory of alliances, suggesting that access to complementary and additional resources is the basic reason for entering into alliances.For example, firms can have access to tangible resources such as financing and investments.In fact, investment sharing by cooperative relationships can reduce costs and be a less uncertain option compared to investing alone.

Market position and product development (factor 5)
This factor had high positive loading on the strategic objectives "product diversification" and "consolidate market position".These variables reflect the nature of product development.Preverez (2001) suggests that product development is a successful outcome in cooperative relationships.Cooperative relationships can create competitive strengths such as vertical linkages (Harrigan, 1985).In fact, these relationships among firms allow them to consolidate their existing market positions.This factor includes variables associated with high uncertainty in the current environment.For example, when the development cost of new products is high, the collaboration instrument minimises the risk of potential failure.

Economies of scale and product rationalisation (factor 6)
The most important objective that includes this factor is economies of scale.Economies of scale may be a way to reduce risk by more efficient cost structures.The use of alliances is also a means of reducing costs; specifically the transaction costs involved in extending vertical links.
In order to validate hypothesis 2 (family businesses have different objectives from non-family businesses), we compared the different means of the factors between the independent samples.The results are shown in Table 6.
There are only a few statistically significant differences between family and non-family businesses regarding objectives for entering into cooperative relationship.Only factor 6 (Economies of scale and product rationalisation) shows a coherent picture.Here, the objectives "to increase production capacity" and "to create economies of scale" both appear to be particularly and consistently important for the family businesses in our sample.Within the other factors, "to improve the level of innovation" of factor 1 and "to consolidate market position" of factor 5 are also statistically relevant objectives for family businesses.
On the other hand, non-family businesses seem to target significantly more technology transfer objectives within factor 1 as well as creating and exploring synergies within factor 3. Despite of the differences identified, given the number of factors with no statistically robust outcomes, there is at best weak support for our second hypothesis.
In the light of these outcomes, family businesses also use cooperative relationships to manage resource scarcity (Das and Teng, 2000), supposedly due to their imperfect organisational and managerial capabilities.Interestingly, despite the weak significances, we could not confirm that family firms pursue objectives in interfirm cooperations that are related to their specific characteristics and firm philosophies.This holds principally for the generally claimed family-oriented long term objectives (Daily and Dollinger, 1993;Harris et al., 1994;Reid and Adams, 2001;Zahra et al., 2004;Zellweger, 2007), for which we could not find evidence when it comes to the reasons for engaging in cooperative relationships.
Contrary to the general arguments (Donckels and Fröhlich, 1991;Kotey, 2005;Hienerth and Kessler, 2006), most of the few significant objectives for inter-firm cooperations cited by family business owners were growth and expansion oriented.

Performance of cooperation
Furthermore, we asked correspondents to value a series of variables, with the purpose of finding out the most important performance indicators in cooperative relationships and the subsequent application of a factor analysis.Table 7 shows the respective answers.Here, "customer satisfaction", "to achieve cooperative objectives" and "to increase annual sales" are the most important indicators of performance in the cooperation processes.The combination of the variables indicates that market position seems to be an important factor for performance evaluation.Interestingly, variables related to access to new markets were also objectives with relatively high means.This nurtures the belief that access to new markets or new market segments is an important benefit of cooperative relationships.
From the factor analysis, three factors were produced which explain a total of 58.54% of the observed variance, as shown in Table 7. Cronbach"s Alpha suggests good internal consistency of each scale as a whole.Interpretation of each of these factors follows.

Strategic management measures (factor 1)
This factor includes subjective indicators such as "balance of power and management", "partner"s influence in decisions", "control shared by several firm areas" and "technology and R and D", simultaneously with objective measures such as "high profit levels".However, Osland and Cavusgil (1996) state that indicators such as increase in exportation or benefits, in spite of being quantitative, may not reflect sufficiently the quality of the relationship.According to Glaister and Buckley (1996), we also included subjective performance measures to be effective in capturing the level of fulfilment of strategic management goals.

Market and operational measures (factor 2)
This factor includes traditional market measures to evaluate firm performance (growth, exports and sales) and qualitative returns to non-financial stakeholders, such as "customer satisfaction".These performance measures of survival, stability and duration were also referred to in Geringer and Hebert"s (1991) study.As noted by these authors, operational measures of performance are the  Notes:The mean is the average on a scale of 1 (="of no importance") to 5 (="of major importance").KMO Measure of sampling Adequacy = 0.811; Bartlett Test of Sphericity: 364.734; df = 78; p < 0.000.Fi (i=1…3) -Factors; Principal components factor analysis with varimax rotation.
most frequently used in cooperation literature.

Cooperative objectives and cultural aspects (factor 3.
"To achieve cooperative objectives" and "cultural compatibility" are the chief indicators within this factor.Geringer and Hebert (1991) also considered the influence of cultural aspects on performance evaluation.Where there are cultural dissimilarities, this may lead to differrences in partners" perceptions of each other and result in a lower degree of performance evaluation.Finally, Table 8 indicates the results of the statistical analysis regarding our hypothesis 3 (family businesses evaluate the performance of cooperative relationships differently).
When considering a set of variables to measure business performance, the results give a heterogenic picture of performance evaluation by family and nonfamily businesses.On the one hand, there is some significance in "partner"s influence in decisions" and "intensity of technology and/or R and D" of factor 1 being more important for non-family businesses.On the other hand, "customer satisfaction" and "previous history of the relationship" of factor 2 as well as "cultural compatibility" of factor 3 have a relatively higher importance for family businesses.Consequently, for a series of measures in our analysis, hypothesis 3 proved true.
The results underline that performance evaluation and the relative importance of the related indicators are a function of the family element within firms.In accordance with the findings of Gulati and Gargiulo (1999), especially for family businesses we could confirm the relevance of prior ties for cooperative relationships."Soft" measures seem to play a major role for family businesses, also and in particular for evaluating cooperative relationships.Cultural and trust-based aspects are of predominant importance.These insights confirm family businesses" strong preference towards non-economic outcomes (Dunn, 1995;Chua et al., 2003;Chrisman et al., 2004;Craig and Moores, 2005;Gómez-Mejía et al., 2007), which on the grounds of our study is also true for their evaluation of cooperative performance.

CONCLUSIONS AND IMPLICATIONS
In this research, we found important differences between small and medium-sized family and non-family businesses when analyzing difficulties and performance evaluation with regard to cooperative relationships.The specificity of family businesses results, thus, in particular behaviour in strategic alliances.These insights lead to a number of implications and recommendations, not just in academic terms as an under-researched topic, but also in policy terms, that is, to assist small and medium-sized family firms with respect to their cooperative strategies.
As exposed in our theoretical reflections at the beginning, family firms stand for a specific corporate culture, reflected by unique norms, values and languages.From our research, we can confirm that for family businesses entering and being successful in cooperative relationships, trust and loyalty are even more important than for their non-family counterparts.Therefore, family businesses should invest in "open" cooperative strategies.In this sense and in accordance with Dunn (1995), we suggest designing separate policies and programmes for promoting cooperative relationships with other family businesses, because they have important characteristics in common.We think that cooperations between family businesses would increase the probability of success.The work of Swinth and Vinton (1993) point towards a similar conclusion.Thus, a strategy based on cooperative relationships with other, perhaps local, family businesses may therefore imply important benefits for the cooperating family businesses and increase their competitiveness.
Furthermore, our research sparks off interesting possibilities for future investigations and new issues for other scholars to build on.First, we recommend that strategic management researchers consider the particularities of family-owned businesses when analyzing strategic alliances.Secondly, we suggest studying cooperative relationships between family businesses on the one hand, and between family businesses and nonfamily businesses on the other.This would allow deeper insights into the characteristics and requirements of cooperative relationships in the sphere of family businesses.Thirdly, the analysis of the behaviour of larger family businesses in cooperative relationships will contribute to the completion of an integrated theory on strategic alliances and family businesses.Fourthly, as cooperative relationships are a process, their evaluation in terms of difficulties, performance, etc. would be better performed in a longitudinal study through various growth stages.This method would permit following up some inter-firm cooperations and evaluating their real progression.
Finally, it should be noted that our study has a number of limitations.The first deals with the small sample.The vast majority of Portuguese firms, both family and nonfamily owned, are SME.So, the low response rate may be due to the fact that for this type of businesses it is more difficult to reply to emailed questionnaires than for larger companies, especially when the subject is relatively new, as is the case with strategic alliances.Another potential argument is that SME managers would have only a limited amount of time to devote to the questionnaire.Due to these arguments, we could not build up a random sample and so a generalisation should be made cautiously.
Another limitation of this research is the "culture effect".As referred to by Papadakis et al. (1998) in their study of Greek firms, nationally owned firms display a national style of management.It is possible that our results are very specific to the Portuguese context.For these reasons, we suggest further research to detect similarities and disparities with family firms of different sizes and regions.Nevertheless, we hope the insights of our study will inspire other scholars, and the combination of this and future work will surely allow valuable comparisons.

H 3 :
Due to the specific characteristics of family businesses, they evaluate the performance of cooperative relationships with different indicators from those used by non-family businesses.

Table 1 .
Research model.Due to the absence of official databases on cooperative relationships in Portugal, we used several secondary sources: (1) databases of firms that participated in programs financed by European Commission funds;

Table 3 .
Difficulties for formation of cooperative relationships: difficulties ranked by mean measure of importance and principal components factor analysis.

Table 4 .
Difficulties in cooperation process.

Table 5 .
Objectives for the formation of cooperative relationships: objectives ranked by mean measure of importance and principal components factor analysis.

Table 6 .
Objectives in the cooperation process.

Table 7 .
Performance indicators of cooperative relationship formation: performance indicators ranked by mean measure of importance and principal components factor analysis.

Table 8 .
Performance in cooperation process.