New venture teams and acquisition: Team composition matters

ABSTRACT

This study examines the relationship between new venture team characteristics on the acquisition likelihood of the new venture. We find that for a US sample of new ventures: new venture teams with equal numbers of females and males, as well as those with higher average education levels are more likely to be acquired. In addition, our results show that higher diversity in industry experience may decrease the acquisition likelihood of new ventures. This implies that teams consisting of both extremes, highly industry experienced and non-experienced members, tend to have the lowest acquisition likelihood all else being equal. This study adds to the rare empirical evidence regarding the factors associated with the exit of new ventures founded by teams, by providing evidence on the effect of team diversity and human capital level on acquisition likelihood. The results may be of interest to both entrepreneurs and external investors.

KEYWORDS:

• New venture teams
• diversity
• exit
• mergers and acquisitions

Introduction

Many large companies utilize acquisition of new ventures as a way to enter and compete in markets for emerging technologies. The healthy and growing market for acquisitions has resulted in many firms being established by entrepreneurs and their financial backers with the explicit desire of being sold to or merging with larger firms in the near future. From the buyer’s perspective, acquisition is a strategic decision which involves significant resources and uncertain outcomes (Deutsch et al., 2007). From the seller’s perspective, the acquisition exit is also a risky path to chase as it may require taking costly actions to suit the desires of potential buyers (Arora & Nandkumar, 2011).

We argue that new venture team characteristics can affect the acquisition likelihood of the new ventures. Upper echelons theory (UET) suggests that the top management team characteristics can be used to partially predict the organization’s strategic choices and performance outcomes (Hambrick & Mason, 1984). The entrepreneurship literature has applied UET to examine the new venture team characteristics which affect new venture performance (Brinckmann et al., 2011; Chandler et al., 2005; Ensley et al., 2002; Hmieleski et al., 2012; Nielsen & Nielsen, 2013; Steffens et al., 2012). In their meta-analysis, Jin et al. (2017) found significant effects of new venture team characteristics, in terms of aggregated human capital and team diversity, on new venture performance. New venture teams are the main resources for their new ventures, so their characteristics and their abilities are expected to strongly affect their ventures.

Despite large interest in new venture team characteristics and their effect on new venture performance, the relationship between those characteristics and new venture exit has mostly been unexamined. Different new venture team characteristics can be more or less important during venture creation, venture growth, and venture exit (Patzelt et al., 2021). The entrepreneurial exit literature has examined how firm-level factors such as product and process innovation (Cefis & Marsili, 2011), age and size (Cefis & Marsili, 2011; Grilli et al., 2010), cash flow, leverage, and debt (Balcaen et al., 2012), employment growth, and legal form (corporation vs sole proprietorship; Cotei & Farhat, 2018) explain the firms’ exits by mergers and acquisitions. A few studies examined the entrepreneur’s characteristics on the selection of acquisition as an exit strategy (DeTienne & Cardon, 2012; DeTienne et al., 2015) and the likelihood of acquisition as a performance outcome (Wennberg et al., 2010). These studies, however, primarily focused on solo entrepreneurs. The effect of new venture team characteristics on acquisition likelihood has seldom been addressed in this literature. Considering the role of team dynamics and the fact that many new ventures are founded and directed by teams (Aldrich et al., 2004), there is a need to study the impacts of team composition on exit paths, including acquisition. One of the rare exit studies that looks at teams is Cotei and Farhat (2018) which examined several drivers of acquisition in a mixed solo and team sample. Some key team composition characteristics such as team diversity were not included in their list of examined drivers.

In this study, we investigate to what extent/how the composition of the new venture team at the time of founding impacts the acquisition likelihood. We focus on the following diversity characteristics in our study: gender diversity and nationality diversity (demographic characteristics), and diversity of education level and industry experience level (human capital characteristics). Although our main emphasis is on team diversity, we also investigate the effects of the average levels of human capital characteristics. This approach is consistent with Jin et al. (2017), which showed that entrepreneurial teams affect performance both through team diversity and the average level of human capital. In some instances, the average levels, rather than diversity, are plausibly related to team performance.

We base our investigation on empirical data from the Kauffman Firm Survey (KFS), which tracked a representative cohort of firms started in 2004 in the US with a baseline survey and seven roughly annual follow-up surveys. We include all three outcomes of survival, merger and acquisition (M&A), and closure in our analysis to examine whether or not factors increasing the acquisition likelihood affect survival likelihood as well. We find that gender diversity and average education level are positively related to acquisition likelihood. Our results also indicate a negative effect from members born outside the US and industry experience diversity of team members on the acquisition likelihood. We find no relationship between education level diversity and acquisition likelihood.

This study has several contributions to the literature. First, the new venture exit literature addresses the effect of solo entrepreneurs’ characteristics on the exit of new ventures; however, there is little focus on the effect of team characteristics (including diversity) on exit. This study provides evidence on the effect of team diversity and human capital on the acquisition of new ventures. Similar to the solo entrepreneurs, our results show that higher human capital in terms of education level increases the chance of successful exit (acquisition) for new ventures founded by teams.

Second, we contribute to the new venture team research by highlighting the importance of later stage performance dimensions (exit, particularly getting acquired) for new venture teams. Exit is a legitimate component of the entrepreneurial process. More than half of new ventures exit less than five years after their entry (Cefis & Marsili, 2011). However, the new venture team literature has mostly focused on drivers of early stage entrepreneurial success such as venture creation and financial performance. Our results demonstrate a lasting strong effect of new venture team composition on new venture outcomes, including exit through acquisition. We also show that the new venture team characteristics which explain the acquisition outcome have impact through channels beyond the traditional performance measures such as profitability. Our study gives partial support to the interpretation of the meta-analysis conducted by Jin et al. (2017) by providing evidence indicating the positive effect both of human capital (average level of education in our case) and diversity (gender diversity in our case) on entrepreneurial success.

Third, we add to the upper echelons theory literature by highlighting the effect of the seller’s top management team on acquisition likelihood. Studies in the upper echelons literature looking at acquisition have mostly focused on the effect of the buyer’s top management team characteristics on acquisition likelihood. For example, Chen et al. (2016) studied the effect of female representation on a firm’s board on the number of acquisitions in which the firm engages. Cao et al. (2019) examined the effect of nationality and cultural diversity on the boards of large US companies on their cross-border merger and acquisition decision effectiveness. In this perspective, acquirers are the decision makers and sellers implicitly have little effect on the acquisition decision. In our view, acquisition is not an event that “happens to” sellers. Rather, it is one in which they are influential and active participants. We show that specific new venture team characteristics can affect acquisition likelihood.

Fourth, the literature on venture capital financing is clear on the notion that venture capital (VC) investors prefer cash-out outcomes like acquisitions, but few studies look at whether the determinants of raising money through VC investors overlap with the determinants of acquisition likelihood. We know from the venture capital literature that a key criterion for investment decisions is the characteristics of the new venture team, superseding even business related characteristics such as product or technology (Gompers et al., 2020). Our results indicate that this use of team characteristics as a signal by VC investors is justified at least in terms of acquisition likelihood.

Theoretical background

Acquisition is an important decision for both buyers and sellers. Buyers deal with uncertain outcomes post-acquisition and sellers may need to undertake risky investments to fit the requirements of potential buyers (Arora & Nandkumar, 2011). We argue that new venture team characteristics can potentially affect the likelihood of acquisition deals through: (1) their impact on new venture quality (fitting buyer’s requirements) and (2) their impact on perceived information asymmetries through signaling.

New venture team composition can affect the new venture quality

Each new venture is characterized by a quality level, which defines its overall probability of success. Upper echelons theory suggests that the top management team characteristics can be used to partially predict the organization’s strategic choices and firm outcomes (Hambrick & Mason, 1984). Upper echelons theory suggests that the level of impact of top management teams on firm outcomes depends on managerial discretion available. New ventures provide more latitude compared to established firms because new ventures are dealing with the liability of newness and the liability of smallness. Jin et al. (2017) demonstrates that UET is applicable to the new venture context. They showed that new venture team composition in terms of aggregated human capital, as well as team diversity, affects the new venture outcomes. New venture team composition can affect the new venture quality and thus the acquisition likelihood of new ventures by:

Providing resources and increasing information processing capability (or decreasing quality through social categorization)

Different types of diversity may affect new venture quality differently. Van Knippenberg et al. (2004) argued that each dimension of diversity may lead to both information processing benefits and social categorization detriments. Each dimension of diversity should be considered within context to predict the impact of information elaboration, categorization, and intergroup bias. We apply social categorization theory and information processing theory in this study to discuss both beneficial and detrimental effects of diversity of each attribute on new venture quality.

Social categorization theory suggests that members’ perceived similarities of demographics, beliefs, values, and attributes result in attraction among team members and the lumping of people into groups/categories. Individuals are inclined to like and trust within-group members and hence generally favor insiders over outsiders, leading to difficulties in inter-subgroup relations and affective (relational) conflicts (Tajfel, 1974).

Information processing theory suggests that diverse teams tend to have access to broader and non-redundant skills, abilities, experience, and knowledge and have distinct opinions and perspectives on the task. Besides providing a larger pool of resources, diversity improves the quality of decision making by forcing team members to reconcile their conflicting perspectives and thus more thoroughly process information (Van Knippenberg et al., 2004). This perspective maintains that diversity leads to increased cognitive (task related) conflict (Jehn et al., 1999) and higher performance and innovation (Bantel & Jackson, 1989; Beckman et al., 2007; Jehn et al., 1999).

In terms of average level of human capital in new venture teams, human capital is shown to be a significant resource for new venture quality. In their meta-analysis, Unger et al. (2011) showed that individual level human capital affects new venture success. Jin et al. (2017) extended this human capital perspective to the team level and they reported a strong effect of aggregated human capital on new venture success. A new venture team’s human capital is a unique, valuable, and difficult to imitate resource, which can lead to competitive advantages. For example, highly experienced and better-educated entrepreneurs have access to higher resources and information, which can increase the quality of the firm in terms of higher quality of decisions, higher financial performance, and greater innovativeness.

Affecting the level of entrepreneurs’ risk taking

The acquisition exit path can be considered a risky one to pursue for sellers. Increasing acquisition likelihood may involve taking costly actions to fit the requirements of potential acquirers which may also increase the chance of closure (Arora & Nandkumar, 2011). Competitive pressures among start-ups competing to get acquired leads them to take greater risks in pursuit of more radical innovations. Norbäck et al. (2011) found that the most high quality innovations are the most attractive acquisition targets, and the model by Henkel et al. (2015) suggests that firms playing the acquisition game will be driven to produce more and more high-risk innovations. We argue that specific new venture team attributes increase the chance that new ventures will be more inclined to pursue the risky path of acquisition. For example, entrepreneurs with higher human capital have higher opportunity costs, which in turn can increase their risk taking. Arora and Nandkumar (2011) discuss that entrepreneurs with higher human capital are more likely to take risky actions to fit the needs of buyers.

New venture team composition can act as a signal

Team composition can act as a signal of legitimacy for new ventures. Information asymmetries can hinder buyers’ ability to access enough information to choose favorable targets and can also hamper sellers’ ability to express their value to buyers (Ragozzino & Reuer, 2007). This hole in perceptions of value can lead to disagreement between the two sides of the acquisition deal and stop acquisitions from happening (Puranam et al., 2006). Information asymmetries are more substantial when the target is an entrepreneurial venture for two primary reasons (Ragozzino & Reuer, 2007): First, buyers often evaluate sellers based on their intangible resources and growth opportunities rather than from assets in place and an established business model. Second, the entrepreneurial venture may not have had enough time to gain legitimacy in the market because of the liabilities of newness or smallness (Aldrich & Auster, 1986; Stinchcombe, 1965). These concerns can aggravate adverse selection but can be alleviated through credible signals (Busenitz et al., 2005). A significant signal of firm value at acquisition can be new venture team characteristics. This is because buyers, who cannot discover the new venture’s value based on economic disclosure, count on “symbolic signals of competence” (Busenitz et al., 2005; Cohen & Dean, 2005; Sine et al., 2006). For example, Packalen (2007) discusses that new ventures can increase their chance of gaining venture capital by adapting to the expectations of investors in terms of the features that effective teams should hold. Our main aim here is to understand how different aspects of new venture team composition influence the new venture acquisition likelihood through the mechanisms mentioned above.

Gender diversity was investigated because several studies indicate that performance of mixed gender teams is superior to gender homogenous teams. The most relevant to this context, indicate that mixed gender teams take greater risks (Díaz-García et al., 2013) and have superior performance (Kravitz, 2003). Based on studies examining social categorization (how in-group people are treated versus out-group), nationality/race was the strongest attribute for categorization (Harrison et al., 2002). We think the relevance of nationality diversity is stronger than race in the new venture context (although we also control for race) because related characteristics such as language or culture could cause issues with communication, dampening the information processing benefits of diversity. Human capital in entrepreneurial teams has been shown to increase performance of the firm. Van der Sluis et al. (2008) discussed that education level positively affects the entrepreneurs’ success (however measured). In this study, we aim to investigate whether this also holds for the specific exit performance measure of acquisition.

Hypotheses development

Demographic characteristics (gender diversity)

Consistent with information processing theory, gender diverse teams bring different experiences which provide different sources of knowledge and heterogeneity of perspectives, leading to higher quality decision making (Del Carmen Triana et al., 2019; Mojambo et al., 2020; Schwab et al., 2016; Singh et al., 2008; Terjesen et al., 2016). As gender diverse teams try to integrate and settle their diverse perspectives, they participate in in-depth discussions and information evaluation (Dahlin et al., 2005), consideration of different criteria in evaluating alternatives (Park, 1996), and thus generate higher quality solutions (Đặng et al., 2020; Hoffman & Maier, 1961). Gender diversity is advantageous when the team’s activity is unstructured, new, and complex (Kravitz, 2003). In new ventures, which are characterized by high complexity (Ensley et al., 2002), gender diversity can enhance decision quality, which in turn improves new venture quality.

Although social categorization theory suggests that diversity may hurt team dynamics by bringing relational conflict and lower cohesiveness (Quintana-García & Benavides-Velasco, 2016; Richard et al., 2004), previous studies argue that gender diversity does not have an effect on perceptions of diversity, whereas other attributes such as race and nationality increase it (Bell et al., 2011; Harrison et al., 2002). McPherson et al. (2001) found that homophily in gender is not as significant as homophily in race and nationality. They suggest that people become familiar with the opposite gender during their early life, for example, through schooling. Hence, by the time they are adults, they have fewer communication problems and relational conflict issues with the opposite gender.

On balance, the benefits related to information processing of gender diverse teams are likely to outweigh any costs associated with social categorization. Previous studies support a positive effect of mixed gender teams on economic performance (Hassan et al., 2017), creativity and innovation (Dai et al., 2019; Østergaard et al., 2011; Sastre, 2014), and R&D quality (Fenwick & Neal, 2001; Xie et al., 2020). Prior research suggests that mixed-gender management teams also tend to be more risk-taking and proactive (Richard et al., 2004), and favor more radical innovations than those led by a single gender (Díaz-García et al., 2013; Griffin et al., 2021). This makes mixed-gender new venture teams more attractive for buyers.

In addition to the above arguments, which suggest a positive effect of gender diversity on firm quality, there is a signaling/legitimacy effect from gender diversity which can increase acquisition likelihood. Pechersky (2016) discusses that the presence of women leads to higher public transparency (a reduction in information asymmetry between company insiders versus outsiders). Nekhili and Gatfaoui (2013) describe this as an “ability of women to improve quality of financial statements” (p. 245). Gul et al. (2011) found that gender diversity increases stock price informativeness through the mechanism of improved public disclosure. The potential positive effect of gender diversity on new venture quality (through strong information processing capability and higher risk-taking) coupled with increased signaling/legitimacy can result in a higher chance of acquisition.

Hypothesis 1: Gender diversity of teams is positively associated with acquisition likelihood for new ventures.

Demographic characteristics (nationality diversity)

Through the lens of information processing theory, it has been argued that nationality diversity in teams can bring different knowledge, experience, and world views, which can contribute to higher quality new ventures (Fernández-Temprano & Tejerina-Gaite, 2020). Consistent with this view, Nielsen and Nielsen (2013) argue that formal (for example, the political and economic rules) and informal institutions (for example, tacit norms, conventions, and values) of origin countries affect the way that individuals perceive and interpret information, and act upon recognized opportunities and threats.

However, social categorization theory suggests that nationality diversity can hurt team processes and thus negatively affect new venture quality. Members from diverse national origins and cultures have different communication styles and different views on how teamwork should proceed (Dahlin et al., 2005). These differences may lead to communication and cooperation problems which in turn impede the efforts to develop a common goal, which is essential for building a successful venture (Vogel et al., 2014). Language diversity is “one of the clearest distinguishing features” (Chen et al., 2006, p. 680) of nationality diverse teams and can largely hurt communication (Henderson, 2005; Khan & Abdul Subhan, 2019), leading to relational conflict, uncertainty, and mistrust among individuals (Feely & Harzing, 2003; Salloum et al., 2019; Tenzer et al., 2014). Nationality diversity leads to strong social categorization as it is even more salient than race, gender, and other demographic traits (Dahlin et al., 2005). On balance, the costs related to social categorization within nationality diverse teams are likely to outweigh any benefits associated with information processing.

Regarding risk-taking, nationality diverse teams have been found to make more conservative decisions (conservative shift) than nationality homogeneous groups (Watson & Kumar, 1992). They discuss that teams with less cultural diversity have less difficulties in their communication and problem-solving, allowing team unity to surface more quickly, and thus a more collaborative attitude of “let’s try it” may surface. The negative effect of nationality diversity on new venture quality (through strong social categorization and lower risk-taking) can result in a lower chance of acquisition.

Hypothesis 2: Nationality diversity of teams is negatively associated with acquisition likelihood for new ventures.

Human capital (education level)

Higher levels of education improve a team’s ability to gather and combine knowledge and information regarding the task (Van der Sluis et al., 2008). In addition, entrepreneurs with higher levels of education have greater access to resources such as financial (Bates, 1990) and social capital (Arenius & De Clercq, 2005) and thus are more likely to leverage those resources to improve the quality of the new venture. In their meta-analysis, Van der Sluis et al. (2008) found that education level positively affects the entrepreneur’s success (however measured).

Higher education level also increases the preference for harvesting the business. Entrepreneurs with greater education have higher opportunity costs and more job options available to them outside of their current venture (Gimeno et al., 1997). Entrepreneurs with higher opportunity costs are more likely to decide to exit in order to switch to new opportunities. These entrepreneurs are more likely to take actions and make investments to fit the needs of buyers (Arora & Nandkumar, 2011).

In addition, higher education level conveys a signal of legitimacy (Cohen & Dean, 2005; Spence, 1973). Packalen (2007) discusses that highly educated people have greater status, which brings them an aura of legitimacy and more confidence in their capability. Thus, ventures with highly educated owners will be more attractive to buyers. The potential positive effect of higher education level on new venture quality (through providing resources and increasing risk-taking) coupled with increased legitimacy can result in a higher chance of acquisition.

Hypothesis 3: The average education level of teams is positively associated with acquisition likelihood for new ventures.

Consistent with information processing theory, high levels of diversity are beneficial to team outcomes if diverse teams can build on a broader range of perspectives than homogeneous teams. However, having members with different education levels does not enhance the varying perspectives required to improve the outcome (Bell et al., 2011). Instead, the factor that can influence success is the overall education level (Bantel & Jackson, 1989; Cooper et al., 1994; Fairlie & Robb, 2007).

Consistent with social categorization theory, education level diversity can hurt the frequency and quality of member interactions. Team members with similar education levels may be more comfortable discussing subjects both related and unrelated to their job (Zenger & Lawrence, 1989). Amason et al. (2006) found that teams with homogenous education levels perform better in highly novel ventures because more homogeneous teams will find high levels of behavioral integration easier to achieve. They discuss that members within homogenous teams communicate more frequently, easily, and less formally. On balance, the costs related to social categorization from education level diversity are likely to outweigh any benefits associated with information processing. This can lead to lower new venture quality, which may result in lower acquisition likelihood.

Hypothesis 4: Education level diversity of teams is negatively associated with acquisition likelihood for new ventures.

Human capital (industry experience level)

Higher industry experience levels bring better information about customers’ expectations and their needs, and also deep knowledge about industry behaviors (Delmar & Shane, 2006). Moreover, industry-experienced entrepreneurs have access to broader networks (for example, suppliers and distributors) within the industry, which in turn provides support (Cooper et al., 1994). Previous studies have found that new ventures whose teams had more industry experience had higher levels of logged sales (Delmar & Shane, 2006) and higher growth (Friar & Meyer, 2003). Also, entrepreneurs with higher industry experience have higher opportunity costs and are more likely to prefer riskier but more high return exit strategies like acquisitions. Highly experienced teams are more likely to take risky actions and make investments to fit the needs of buyers (Arora & Nandkumar, 2011).

Beside the positive effect of industry experience levels on the new venture quality, industry experience may convey the legitimacy of ventures (Cohen & Dean, 2005). The familiarity with industry behaviors (norms, practices, and routines) of experienced entrepreneurs attracts investors who consider this factor to be a key predictor of new venture success (Packalen, 2007). The potential positive effect of industry experience level on new venture quality (through providing resources and increasing risk-taking) coupled with increased legitimacy can result in a higher chance of acquisition.

Hypothesis 5: The average industry experience of teams is positively associated with acquisition likelihood for new ventures.

A similar theory framework to education would apply to the industry experience diversity of team members. Foo (2011) found that average industry experience positively affects member-rated team effectiveness, but diversity of industry experience has a negative effect on the outcome. Diversity of prior industry experience may lead to conflict between experienced and inexperienced team members and hence decrease team cohesion and effectiveness. Industry experience diversity results in lower trust and lower confidence in the team’s ability among team members, hence decreasing individuals’ willingness to provide resources to their teams (Davis et al., 2009), and results in more turnover in teams (Chandler et al., 2005). On balance, the costs related to industry experience diversity are likely to outweigh any benefits associated with information processing. This can lead to lower new venture quality, which may result in lower acquisition likelihood.

Hypothesis 6: Industry experience diversity of teams is negatively associated with acquisition likelihood of new ventures.

Methods

Data collection

In this study, we used the Kauffman firm survey (KFS) from the Ewing Marion Kauffman Foundation. The KFS is a panel study of 4,928 new ventures that started in 2004 within the US, with follow-up data annually until 2011. The KFS is unique in that it represents a large sample of new ventures with data on firms of varying sizes/industries, and it follows those new ventures from their establishment annually. Furthermore, all firms are born in the same period (in 2004), hence controlling for age, economic cycles, and other time-varying noise. For our specific purpose of studying survival and acquisition, the KFS data has the benefit of including a large sample in which a significant number of firms failed or were acquired. “The venture exit is a complex process to empirically measure because it requires gathering data by asking business owners after exit to share information on the type of exit” (Cotei & Farhat, 2018, p. 545).

The longitudinal panel-form of the dataset only includes new ventures that responded to all surveys up to the last year they existed. “Sample attrition, refusals, unlocatables, and firm closures” lead to sample loss in each year (Robb et al., 2009). This results in a baseline sample of 3,140 firms, including 900 team-founded ventures. By design, female-owned and technology firms were intentionally oversampled. Sampling weights in the data correct for both over-sampling and survivorship bias (Litwin & Phan, 2013). Out of 900 team-founded new ventures in the initial sample, 462 survive until 2011 (73 had exited through M&A and 365 had closed).

Measures

Definition of the variables used in our analysis and their measurements are listed in Table 1.

Table 1. Variable definitions.

	Variable	Definition
Dependent Variable	Acquisition	1 = M&A, otherwise 0
	Closure	1 = Closure, otherwise 0
Independent Variables	Gender diversity	Standard deviation = square root of (p*(1-p)), where p is the proportion of female in the first year
	Nationality diversity	Standard deviation = square root of (p*(1-p)), where p is the proportion of non-US born in the first year
	Average education level^a	Average of education level per firm founder in the first year
	Education level diversity	Blau’s index of education level per firm founder in the first year
	Average industry experience	Average number of years of work experience in the industry per firm founder in the first year
	Industry experience diversity	Standard deviation of number of years of work experience in the industry per firm founder in the first year
Control Variables	Average age	Average reported age per firm founder in the first year
	Age diversity	Standard deviation of reported age per firm founder in the first year
	Race diversity	Standard deviation = square root of (p*(1-p)), where p is the proportion of non-white in the first year
	Commitment – hours worked	Average reported weekly working hours per firm founder
	Commitment – equity owned	Percentage of equity owned by founders in the first year
	Couple	1 if the size of team is equal to two and the spouse own the equity in the firm, otherwise 0
	Team size	Number of active owners in the first year
	Industry technology level^b	1 if firm is in high-tech category, 0 otherwise
	Firm size	Logarithm of total assets of the firm in the first year
	Patent	1 if the venture has patent(s) in the first year, otherwise 0

^aCategories: 1 = less than 9th grade; 2 = some high school; 3 = high school diploma/equivalent; 4 = technical/trade/vocational degree; 5 = some college; 6 = associate’s degree; 7 = bachelor’s degree; 8 = some graduate school; 9 = master’s degree; and 10 = professional school or doctorate.

^bBased on the percentage of employment dedicated to research and development in the firm, the KFS determined four industries as high tech: chemicals and allied products, industrial machinery and equipment, electrical and electronic equipment, and instruments and related products.

Dependent variables

The primary dependent variable is acquisition likelihood. We define an event variable which takes the value 1 if the firm was merged or acquired that year, and 0 otherwise. We eliminate acquisitions where revenue was non-positive and the firm did not meet growth expectations (or unknown). This is done to alleviate concerns that distress sales are lumped in with successful acquisition exits. Entrepreneurs were asked in the 2008 survey: “How much do you think your business met your expectations for growth between when the business was started and December 31, 2008?” Entrepreneurs could respond: (1) exceeded; (2) met; or (3) did not meet. These measures eliminate 11 acquisitions. Arora and Nandkumar (2011) and Eesley et al. (2014) applied similar filtering criteria to measure favorable acquisitions. Also, we test the effect of our independent variables on closure likelihood to check if targeting acquisition may lead to higher risk of closure. To do so, we define a variable for closure likelihood. This variable takes a value of 1 if the firm closed, and 0 otherwise.

Independent variables (team characteristics)

We measure diversity in nonbinary categorical variables using Blau’s index, calculated as $\mathit{B}=\left[1-\sum p_i\right]$, where p is the ratio of team members in a category i. This index is often applied to measure diversity of categorical variables (Chowdhury, 2005; Jin et al., 2017).

In our study, diversity of binary categorical variables is calculated as the standard deviation = square root of (p*(1-p)), where p is the proportion of team members in a category. For binary variables standard deviation is less biased than Blau’s index for small group sizes (Biemann & Kearney, 2010). Results are robust to using Blau’s index for diversity in the binary categorical variables.

For diversity of numerical variables, the coefficient of variation, which is the standard deviation divided by the mean, is sometimes used in the literature (Foo, 2011). Because we also include the mean of all the numerical variables in our analysis, we measure the diversity of the numerical variables using the standard deviation.

Demographic characteristics (gender and nationality)

Gender and nationality diversity are calculated by standard deviation. Although our main emphasis is on team diversity, we also investigate the effects of the proportion (female, non-US born) on team performance. In some instances, the proportion of specific groups, rather than diversity, are more plausibly related to team performance.

Human capital (education and industry-specific work experience)

Education of members ranged from fewer than nine years to a PhD degree. Education diversity was calculated with Blau’s index. We also include the average education level of the team, which is the mean of numerical responses of all founders of the firm in 2004.

For an industry-specific experience variable we calculated diversity in the number of years of experience using standard deviation. We also include the average industry experience of the team, which is the mean of numerical responses of all founders of the firm in 2004.

Control variables

We use a number of controls specific to the issue of team composition. We control race diversity to distinguish between the effects of nationality and race. Each of these two terms has a particular definition and links to specific theoretical backgrounds. We argue that individuals from different races but born in the same country are likely to experience lower miscommunication and affective (relational) conflict compared to when they come from different countries. The literature unfortunately often uses these two terms as interchangeable predictors for entrepreneurial activities and does not distinguish between them (Chandler et al., 2005; Foo, 2011; Hellerstedt et al., 2007; Timmerman, 2000). In addition, we control for average age and age diversity of teams to distinguish between the effects of industry experience and age. Ruef et al. (2003) demonstrates that romantic teams are a different category; hence, we use a dummy as a control for couple teams. Team size is controlled for because team size has been shown to affect acquisition likelihood (Soleimani & Keyhani, 2021) and team diversity measures are sensitive to team size (Steffens et al., 2012). The level of commitment has also been controlled for. Individuals’ commitment to their business can affect their acquisition likelihood. The extent of owners’ work can demonstrate their commitment to their business and may signal the legitimacy of the business to buyers. We define two variables to include the effect of owners’ commitment: (1) founders work hours measured in weekly hours allocated, and (2) total equity owned by founders (percentage). We control for firm size measured by the logarithm of total assets in the first year. Firm size is a factor that affects acquisition and survival of firms (Wennberg et al., 2010). We control for whether founders have a patent at venture inception. Patents may signal the quality of an innovative idea to investors or potential buyers. Patents earned later are not controlled for as they are a performance outcome of the venture. Industry technology level also is controlled for since the technological environment impacts the exit strategy of a firm. For instance, the newspaper publishing industry has a lower rate of M&A compared to other industries (Cefis & Marsili, 2011). Finally, to account for temporal dependence, year dummies were included in all models.

Analysis and results

Descriptive statistics

The pairwise correlation coefficients and the summary statistics (means and standard deviations) for all variables are reported in Table 2. The correlation statistics in Table 2 show that team size is correlated positively with acquisition likelihood. Team size is also correlated positively with nationality diversity, race diversity, industry experience diversity, age diversity, and education diversity. This indicates the sensitivity of team diversity to team size. To check multicollinearity, the variance inflation factor (VIF) was computed for all of the independent variables. The results show that there is no serious problem of multicollinearity, because all our VIF levels are in the range of 1.01 to 1.45 (below 10).

Table 2. Descriptive statistics and correlations.

	Variable	1	2	3	4	5	6	7	8		9	10	11	12	13	14	15	16	17	18
1	Acquisition	1.000
2	Closure	−0.030*	1.000
3	Gender diversity	−0.001	0.002	1.000
4	Couple	−0.011	0.013	0.006	1.000
5	Nationality diversity	−0.008	−0.022	−0.110*	−0.024	1.000
6	Race diversity	0.005	0.009	0.006	−0.038*	0.080*	1.000
7	Average education	0.027	−0.050*	−0.086*	0.031*	0.220*	0.049*	1.000
8	Education diversity	0.010	0.018	0.087*	−0.040	−0.093*	0.021	−0.110		1.000
9	Average experience	−0.004	−0.049*	−0.173*	−0.073*	0.035*	0.036*	0.028*		−0.035*	1.000
10	Experience diversity	−0.028*	−0.025	0.065*	−0.033*	0.032*	−0.025	−0.082*		0.103*	0.409*	1.000
11	Work hours	0.028*	−0.053*	−0.165*	−0.011	0.071*	0.059*	−0.072*		−0.035*	0.092*	0.034*	1.000
12	Equity owned	−0.030*	−0.011	0.164*	−0.020	−0.136*	−0.042*	−0.139*		−0.030*	−0.075*	−0.068*	−0.051*	1.000
13	Average age	−0.005	0.009	0.089*	−0.020	−0.022	0.001	0.124*		−0.020	0.345*	0.228*	−0.136*	−0.041*	1.000
14	Age diversity	−0.001	−0.011	−0.259*	−0.060*	−0.004	0.021	−0.022		0.183*	0.099*	0.208*	0.029*	−0.034*	0.087*	1.000
15	Team size	0.049*	−0.030*	−0.073*	−0.082*	0.013	0.102*	0.063*		0.215*	0.059*	0.079*	−0.013	−0.147*	−0.008	0.184*	1.000
16	Firm size	0.037*	−0.058*	−0.046*	−0.018	0.059*	−0.022	0.002		0.007	0.024	0.068*	0.175*	−0.116*	0.076*	0.024	0.082*	1.000
17	Technology level	−0.007	−0.041*	−0.120*	−0.014	0.081*	0.059*	0.059*		0.034	0.159*	0.113*	0.094*	−0.146*	0.099*	0.088*	0.065*	0.075*	1.000
18	Patent	0.013	−0.022	−0.104*	0.012	0.165*	0.006	0.220*		−0.066*	0.079	0.045	0.031	−0.192*	0.063*	0.075*	0.019	0.060*	0.252*	1.000
Mean		0.012	0.057	0.249	0.034	0.032	0.051	6.406		0.350	11.461	6.303	36.710	94.696	45.455	5.412	2.281	9.761	1.185	0.065
SD		0.110	0.232	0.247	0.181	0.121	0.149	1.801		0.239	8.317	6.157	19.214	16.551	9.492	5.365	0.648	3.602	0.388	0.246

 *p < .05.  

The effect of new venture team composition on the acquisition likelihood

In this study, the dependent variable is an event; therefore, survival analysis is suitable since it accounts for right-censoring. Logistic regression is used for semi-parametric survival analysis when event times are discrete.

Tables 3 and 4 respectively present the results of the binary logit regressions predicting acquisition and closure likelihood. Odds ratios are reported in Tables 3 and 4. Odds ratios show the likelihood of one exit type (acquisition or closure) relative to the likelihood that specific exit does not happen. Therefore, coefficients greater than one suggest the variable increases the acquisition or closure likelihood, and coefficients less than one suggest the variable decreases the acquisition or closure likelihood. Proceeding further, whenever we use the term “likelihood,” we are referring to odds ratios, unless otherwise stated.

Table 3. Logistic regression for acquisition.

	Model 1	Model 2	Model 3	Model 4
Controls
High-tech	1.056 (0.371)	1.177 (0.392)	1.121 (0.396)	1.166 (0.404)
Firm size	1.088^+ (0.048)	1.105* (0.050)	1.115* (0.052)	1.061 (0.045)
Patent	0.655 (0.654)	0.729 (0.633)	0.764 (0.645)	0.885 (0.657)
Average age	0.988 (0.017)	0.978 (0.018)	0.977 (0.019)	0.977 (0.020)
Age diversity	0.975 (0.030)	1.013 (0.029)	1.004 (0.031)	1.008 (0.031)
Work hours	1.009 (0.008)	1.014 (0.008)	1.012 (0.009)	1.008 (0.009)
Equity owned	0.989 (0.008)	0.983* (0.007)	0.984* (0.007)	0.978** (0.007)
Team size	1.278* (0.097)	1.305** (0.083)	1.329** (0.087)	1.354*** (0.081)
Couple	0.085* (1.036)	0.070* (1.052)	0.061** (1.057)	0.085* (1.056)
Race diversity	0.837 (1.202)	1.027 (1.120)	1.023 (1.140)	1.372 (1.143)
Profitability				1.001* (0.000)
Independent variables
Gender diversity		4.931* (0.658)		5.864* (0.690)
Nationality diversity		0.087 (1.620)		0.110 (1.652)
Average education level		1.253* (0.093)	1.262* (0.094)	1.244* (0.099)
Education level diversity		0.884 (0.698)	1.103 (0.709)	0.966 (0.720)
Average industry experience		1.022 (0.021)	1.018 (0.021)	1.017 (0.022)
Diversity industry experience		0.898** (0.035)	0.902** (0.038)	0.922* (0.033)
Female ratio			2.711* (0.415)
Non-US born ratio			0.062^+ (1.590)
Constant	0.008** (1.508)	0.003*** (1.622)	0.003*** (1.636)	0.003*** (1.65)
N	4,537	4,508	4,508	4,214
F	6.84	6.31	7.25	6.81
Prob > F	0.000	0.000	0.000	0.000

⁺p <.10, *p <.05, **p <.01, ***p <.001.

Standard errors in parentheses.

Table 4. Logistic regression for closure.

	Model 1	Model 2	Model 3	Model 4
Controls
High-tech	0.663* (0.209)	0.704^+ (0.213)	0.695^+ (0.211)	0.747 (0.222)
Firm size	0.972 (0.018)	0.976 (0.019)	0.975 (0.019)	0.980 (0.019)
Patent	0.742 (0.517)	0.944 (0.497)	0.928 (0.533)	0.735 (0.579)
Average age	0.997 (0.008)	1.005 (0.008)	1.005 (0.008)	1.006 (0.008)
Age diversity	0.999 (0.013)	0.987 (0.014)	0.993 (0.013)	0.980 (0.015)
Work hours	0.992* (0.004)	0.991* (0.004)	0.992* (0.004)	0.991* (0.004)
Equity owned	0.999 (0.005)	0.997 (0.005)	0.997 (0.005)	1.001 (0.005)
Team size	0.799^+ (0.122)	0.751* (0.139)	0.736* (0.141)	0.691** (0.138)
Couple	1.601 (0.344)	1.720 (0.351)	1.864^+ (0.350)	1.435 (0.368)
Race diversity	0.937 (0.466)	0.939 (0.489)	0.903 (0.480)	0.993 (0.520)
Profitability				0.999** (0.000)
Independent variables
Gender diversity		0.550+ (0.333)		0.535+ (0.341)
Nationality diversity		0.623 (0.835)		0.498 (0.919)
Average education level		0.872** (0.044)	0.869** (0.044)	0.857** (0.047)
Education level diversity		2.352** (0.331)	2.253* (0.320)	3.540*** (0.364)
Average industry experience		0.980^+ (0.011)	0.981^+ (0.011)	0.980^+ (0.011)
Diversity industry experience		0.990 (0.014)	0.988 (0.013)	0.992 (0.014)
Female ratio			0.637 (0.281)
Non-US born ratio			1.085 (0.527)
Constant	1.065 (0.709)	1.068 (0.780)	1.064 (0.780)	1.064 (0.847)
N	4,537	4,508	4,508	4,214
F	3.36	3.44	3.42	3.34
Prob > F	0.000	0.000	0.000	0.000

⁺p <.10, *p <.05, **p <.01, ***p <.001.

Standard errors in parentheses.

Model 1 includes the controls, whereas Model 2 adds the effects of team composition, and Model 3 implements the proportions of female and non-US born. To provide evidence that the acquisition outcome is empirically distinct from conventional measures of performance, we model the acquisition likelihood of the new businesses as a function of profitability and the independent and control variables in Model 4. If the acquisition likelihood is different from other conventional performance measures (profitability), we would expect some independent variables to still have significant predictive power.

Hypothesis 1 predicted a positive effect of gender diversity on acquisition likelihood. Model 2 strongly supports hypothesis 1 and indicates a positive relationship between gender diversity and acquisition likelihood (OR = 4.931, p < .05) and shows weak support for a negative effect of gender diversity on closure likelihood (OR = 0.550, p < .1). So, a team with maximum gender diversity (equal number of males and females) is 2.2 times more likely to be acquired than a gender homogenous team. Also, in Model 3, we can observe that the proportion of women on the team is positively and significantly related to acquisition likelihood (OR = 2.711, p < .05). The size of this effect is smaller that of gender diversity.

Hypothesis 2 predicted a negative effect of nationality diversity on acquisition likelihood. Model 2 does not support the negative effect of nationality diversity on acquisition (OR = 0.087, p > .10). However, Model 3 shows weak support for a negative effect of non-US born proportion on acquisition likelihood (OR = 0.062, p < .1). Our results indicate that a higher proportion of US-born members on the team improves acquisition likelihood, implying that cooperating with US-born can be a viable way for individuals who are born outside the US to gain access to resources that they could not provide themselves. Non-US born could be disadvantaged by their unfamiliarity with the US market (for example, fewer contacts and greater difficulty in assembling resources; Cooper et al., 1994). In addition, Model 2 does not show any effect of nationality diversity on closure likelihood (OR = 0.623, p > .10).

Hypotheses 3 and 4 predicted a positive effect of average education level and a negative effect of education level diversity on acquisition likelihood respectively. Model 2 strongly supports Hypothesis 3 and indicates a positive relationship between average education level and acquisition likelihood (OR = 1.253, p < .05). This indicates that, all else being equal, a team where all members hold a PhD is 6.1 times more likely to be acquired than a team with all members having fewer than nine years of education. However, results show no effect of education level diversity on acquisition likelihood (OR = 0.884, p > .10). Therefore, Hypothesis 4 is not supported. In addition, Model 2 shows that average education level decreases closure likelihood (OR = 0.872, p < .01), but education level diversity increases it (OR = 2.352, p < .05). These results indicate that higher education level benefits the successful exit of the new venture while reducing closure likelihood.

Hypotheses 5 and 6 predicted a positive effect of average industry experience and a negative effect of industry experience diversity on acquisition likelihood respectively. No effect of the average industry-specific experience is found (OR = 1.022, p > .10). Therefore, Hypothesis 5 is not supported. Our results from Model 2 indicate a negative effect of diversity of industry-specific experience on acquisition (OR = 0.898, p < .01). Therefore, Hypothesis 5 is not supported but Hypothesis 6 is supported.

In results not reported, we also tested the squared term and square root of our independent variables in our regressions to test for curvilinear effects. The results were not significant.

Looking at Model 4 which controls for profitability, we see that the significance of all results above still hold. This indicates that there is an effect on acquisition in addition to the channel through traditional performance. In total, although acquisition likelihood is highly correlated with traditional financial performance, it is a unique performance measure.

Discussion

Consistent with upper echelons theory (Hambrick & Mason, 1984), the new venture team literature shows a significant effect of new venture team characteristics in terms of aggregated human capital and team diversity on new venture performance. In their meta-analysis, Jin et al. (2017) argue that the effects of new venture team characteristics on new venture performance may be stronger than the effect of top management team characteristics in the context of established firms. However, this literature neglects the effect of new venture team composition on later stage measures of performance (exit, particularly acquisition; Wennberg & DeTienne, 2014). Different new venture team characteristics can be more or less important during venture creation, venture growth, and venture exit (Patzelt et al., 2021). Our findings indicate that several demographic and human capital attributes of a new venture team matter with regards to acquisition likelihood. This demonstrates a lasting strong effect of new venture team composition on new venture success. We found that gender diversity affects acquisition likelihood positively, while industry experience diversity decreases acquisition likelihood, and nationality diversity does not affect acquisition likelihood. These results are not fully consistent with the result by Jin et al. (2017) who found a positive impact of entrepreneurial team diversity on new venture success. The reason can be that in their meta-analysis they merge all measures of diversity together, including diversity in education background, functional experience, gender, age, education level. Van Knippenberg et al. (2004) discuss that no specific characteristics are necessarily positively or negatively related to team performance. Each dimension of diversity should be considered within context to predict the impact of the contingencies of elaboration, categorization, and intergroup bias.

In our study, negative team dynamic effects from social categorization associated with gender diversity are outweighed by the information processing benefits of the gender diversity for a net improvement to new venture success (acquisition here). However, for nationality diversity, the negative and positive effects approximately balance each other for no net effect. This may be due to less pronounced social categorization resulting from gender than that of other demographic characteristics such as nationality (Harrison et al., 2002; McPherson et al., 2001). In addition, women in teams have been shown to promote transparency (Pechersky, 2016). Given the popularity of the topic of women in top management teams both within academia and in the popular press, it is likely that investors see presence of women in top management teams to be beneficial for the firm (Welbourne et al., 2007). Vogel et al. (2014) showed that potential investors consider teams that are mixed in terms of gender to be more promising, and they are more willing to support their businesses.

While we found higher diversity of human capital (in terms of industry experience) decreases acquisition likelihood, our study shows that higher average human capital (in terms of average education level) increases acquisition likelihood and decreases closure likelihood. Benefits from human capital arise because higher human capital can lead to (1) higher quality new ventures (by providing resources and increasing risk-taking) and (2) increased legitimacy through signaling. Detienne and Cardon (2006) and Wennberg et al. (2010) found a positive effect of solo entrepreneur’s human capital on successful exits. We extend this result to the new venture team context in terms of acquisition exits. Higher human capital diversity may hurt new venture outcomes by introducing communication and relational problems, while not providing significant information processing advantages (Amason et al., 2006; Bell et al., 2011). Teams consisting of both extremes, highly experienced business founders and non-experienced members, tend to have the lowest success (Foo, 2011). This result implies that homogeneity in terms of industry experience level can lead to higher success by easing the communication among members.

This paper contributes to the existing literature on the exit of new ventures. Our study adds to the rare empirical evidence regarding the factors associated with the exit of new venture founded by teams, by providing evidence on the effect of team diversity and human capital level on the acquisition of new ventures. We showed that acquisition as a performance measure is different from frequently used traditional performance measures (for example, profitability). Our analysis shows that team characteristics still affect acquisition likelihood when profitability is controlled for. This emphasizes the need to include exits such as acquisition as a success measure of performance.

Studies in the upper echelons literature looking at acquisition have mostly focused on the buyer-side perspective (Graebner & Eisenhardt, 2004). In that perspective, the acquirer has been the focus as the decision maker of importance and the seller implicitly has little effect over the acquisition decision. We contribute to this literature by considering the effect of the characteristics of seller’s management on acquisition likelihood. We find that the composition of a new venture team does indeed matter with regard to acquisition likelihood of new ventures.

This work has implications for entrepreneurs and venture capitalists. Investors, including venture capitalists, consider acquisition as a highly favorable exit through which they can harvest their investments. However, information asymmetries are pronounced in new ventures. Management teams can provide credible signals about the prospects of a venture to attract investors. Venture capitalists and other investors have been utilizing management teams and their characteristics as a signal in their investment selection, superseding even business-related characteristics such as product or technology (Gompers et al., 2020). Our results indicate that this use of teams’ composition as a signal by investors is justified at least in terms of acquisition.

At the far end of the spectrum, serial entrepreneurs always have their eyes on the next opportunity. They are often looking to sell their current ventures to be able to free up capital to invest in their next new venture. A contribution of this research for entrepreneurs is to indicate that specific team compositions such as higher gender diversity and higher average education level helps to increase the chance of acquisition.

Limitations and future research

This study has several limitations which, in turn, provide opportunities for future research. First, there are limitations due to the data available from the KFS. There are important attributes such as functional background and education major, which may affect acquisition likelihood. It would be expected under UET that there would be information processing benefits to diversity in these characteristics. However, the KFS does not provide data on these variables. Regarding the nation of origin, there are other correlated factors which may be driving the effect, such as the place of their upbringing, citizenship, or language. Future research may explore the channels of causation from these more fine-grained measures.

Second, as has been argued by Van Knippenberg et al. (2004), the context in which top management team heterogeneity is examined is important. Simons et al. (2017) emphasizes that all studies are constrained by their context as to the extent the results are generalizable. The specific context we examined is the United States. In fact, our results are not in line with Nielsen and Nielsen (2013) who found a positive effect of nationality diversity on economic performance in Switzerland. Switzerland has the highest percentages of foreign executives in Europe and also Swiss firms have a relatively high degree of internationalization due to the small home market size and long history of international operations. In the context of the US, which is a relatively self-contained market (say measured by trade to GDP), building nationality diverse teams may not be significantly helpful for US-born entrepreneurs. In addition, it is possible that the benefits of gender diversity would be diminished in a country with a more negative attitude toward women in the workplace and/or one with more segregation of genders during formative years. Future research may benefit from examining other countries or conducting cross-country studies. Some exit routes can have different interpretations in terms of venture success in different markets. Acquisition is procyclical in the US but countercyclical in Japan, because acquisitions are linked to “rescue mergers” in Japan (Cotei & Farhat, 2018).

Third, because all of the businesses in our study were founded in 2004, they were in a key stage of development when the financial crisis of 2008 hit. Financial/economic crises surge the likelihood of firms’ exit through liquidation. As a result, it is likely that closures were higher and acquisitions lower than if the sample was in a more “normal” period. Although the KFS is a valuable database for studying acquisition likelihood of new ventures, a new survey would be valuable to study exits further. It would be beneficial in the future to look at businesses founded post-2008 crisis. Although, the COVID crisis will pose similar issues for the next longitudinal surveys and studies. We have no reason to believe that our quantitative results are driven primarily by “other characteristics of the participants, materials, or context.”

Conclusion

We examine to what extent/how the composition of the new venture team impacts the acquisition likelihood. We focus on team level characteristics, specifically team diversity, as factors influencing acquisition likelihood. We argue that team compositions can affect the acquisition likelihood of new ventures through (1) its impact on new venture quality (fitting buyer’s requirements);, and (2) its impact on perceived information asymmetries through signaling.

We find that new venture teams with equal numbers of females and males, as well as those with higher average education levels are more likely to be acquired. Also, we find that higher diversity in industry experience may decrease the likelihood of new venture teams being acquired. This implies that teams consisting of both extremes, highly experienced and non-experienced members, tend to have the lowest acquisition likelihood. We add to the entrepreneurial exit literature by analyzing the determinants of acquisition exit of entrepreneurial firms by focusing on new ventures team characteristics. The results may be of interest to both founders and external investors.

Disclosure statement

No potential conflict of interest was reported by the authors.