• Walang Nahanap Na Mga Resulta

Determinants of Entrepreneurial Venture Growth in the Philippines Using the

N/A
N/A
Protected

Academic year: 2023

Share "Determinants of Entrepreneurial Venture Growth in the Philippines Using the "

Copied!
72
0
0

Buong text

(1)

De La Salle University De La Salle University

Animo Repository Animo Repository

Angelo King Institute for Economic and

Business Studies Units

2-2021

Determinants of Entrepreneurial Venture Growth in the Philippines Determinants of Entrepreneurial Venture Growth in the Philippines Using the Global Entrepreneurship Monitor

Using the Global Entrepreneurship Monitor

Brian C. Gozun John Paolo R. Rivera

Follow this and additional works at: https://animorepository.dlsu.edu.ph/res_aki

Part of the Entrepreneurial and Small Business Operations Commons, and the International Business Commons

(2)

Determinants of Entrepreneurial Venture Growth in the Philippines Using the

Global Entrepreneurship Monitor

DLSU-AKI Working Paper Series 2021-02-062 By: Brian C. Gozun (+)

De La Salle University John Paolo R. Rivera

Asian Institute of Management

(3)

Determinants of Entrepreneurial Venture Growth in the Philippines Using the Global Entrepreneurship Monitor

Brian C. Gozun

De La Salle University, Manila, Philippines brian.gozun@dlsu.edu.ph

John Paolo R. Rivera

Asian Institute of Management, Makati City, Philippines jrivera@aim.edu

Abstract

Entrepreneurial venture growth requires the capacity to produce products that are acceptable to the market, and the level of support given to enterprises helps them produce, innovate, and gain market access. However, entrepreneurs are faced with challenges related to physical and social infrastructure, local and global business environment, a level playing field, access to financing, and access to skill development and knowledge. If these remain unmitigated, they have the potential to hamper entrepreneurial growth. Hence, we inquire on the critical drivers of venture growth that will allow entrepreneurs to stimulate their enterprises using founder characteristics, firm attributes, and entrepreneurial strategies. Using the Global Entrepreneurship Monitor (GEM) data for the Philippines, we found empirical evidence that entrepreneurial strategies are being moderated by founder characteristics and form attributes in driving entrepreneurial venture growth. We recommended interventions that will enable enterprises to increase their international orientation and export participation through enhanced access to global value chains.

JEL Classification: C13, L21, L26

Keywords: entrepreneurship, entrepreneurial venture growth, and global entrepreneurship monitor

(4)

Introduction

Technical skills, business acumen, and flexibility drive entrepreneurial competitiveness, entrepreneurial venture growth, and eventual entrepreneurial success in the midst of an ever- changing business climate due to globalization (Sykes, 2017; Sanyang & Huang, 2010; Parker, 2009). For Cooney (2012), the skillsets required to be an entrepreneur are classified into three groups, namely, “entrepreneurship skills, technical skills, and management skills” (p. 7). Similarly, Storey (1994) cited three groups of determinants of entrepreneurial venture growth: founder characteristics, firm attributes, and strategies of the entrepreneur. Likewise, entrepreneurial success is also determined by the strength of the entrepreneurial ecosystem (Bischoff, 2019) or the entrepreneurial economy (Drucker, 1984). Hence, the success of entrepreneurial ventures promotes economic competitiveness (Kritikos, 2014), growth, and development (Moscoso, 2017).

In the Philippines, entrepreneurship1 is an important engine of economic growth that can empower the poor, enhance production, and stimulate innovation (Evangelista, 2013). In 2019, according to the List of Establishments of the Philippine Statistics Authority (PSA), as reported by the Department of Trade and Industry [DTI] (2020), there were a total of 1,000,506 business enterprises operating in the country (from 915,726 in 2016) of which 202,011 (20.2%) are situated in Metro Manila.2 Micro, small, and medium enterprises (MSMEs3) account for 995,745 (99.57%)

1 In this study, we follow the definition of entrepreneurship by the Global Entrepreneurship Monitor (GEM) – "any attempt at new business or new venture creation, such as self-employment, a new business organization, or the expansion of an existing business, by an individual, a team of individuals, or an established business"

(http://gemconsortium.org/wiki/1149).

2 It is important to understand that Metro Manila and Manila are two different places. Following Gaerlan (2015), Manila, whose complete name is the City of Manila, is the official capital of the Philippines. Meanwhile, Metro Manila is the region (i.e., National Capital Region or NCR) where the City of Manila is located, together with 15 other cities (i.e., Caloocan, Las Piñas, Makati, Malabon, Mandaluyong, Marikina, Muntinlupa, Navotas, Parañaque, Pasay, Pasig, Quezon City, San Juan, Taguig, and Valenzuela) and one municipality (i.e., Pateros).

3 The Philippines uses two bases in operationally defining MSMEs – employment and asset size. The PSA uses employment while the Small and Medium Enterprise Development Council (SMEDC) uses asset size as basis for classification. For the specific brackets for employment and asset size, see https://dtiwebfiles.s3-ap-southeast- 1.amazonaws.com/BSMED/MSME+2019+Statistics/2019+Philippine+MSME+Statistics+in+Brief.pdf

(5)

of the total establishments, of which 891,044 (89%) were micro enterprises, 99,936 (10%) were small enterprises, and 4,765 (0.5%) were medium enterprises. Large enterprises made up the remaining 4,761 (0.5%). These MSMEs (83.85%) were from the following industries: wholesale and retail trade, repair of motor vehicles and motorcycles, accommodation and food service activities, manufacturing, other service activities, and financial and insurance activities. Also, MSMEs generated a total of 5,510,760 (62.4%) of the country’s total employment in 2019.

Although MSMEs in the Philippines have always been in the pursuit of overcoming constraints “such as access to finance, technology and skills; information gaps; and difficulties with product quality and marketing” (de Vera, 2012, p. 350), they have been resilient in continuously creating employment opportunities, contributing to economic value-added, and figuring prominently in export trade (Bolido, 2020). With the coronavirus pandemic impacting this backbone of the Philippine economy, Carlos (2020) reported that “in the spirit of Filipino entrepreneurship, many businesses are using this time to strategize” (par. 6) by pivoting to improve short term operations while weathering the pandemic (e.g., online selling, remote staffing, shift to production of essential goods). In realizing the full growth potential of MSMEs, Fong (2018) discussed that they should be ready and open to internal learning; adopt entrepreneurial activities to adapt and succeed in a highly competitive business ecosystem; and flexible in redeploying resources and recalibrating goals—all of which can drive innovation and success.

Entrepreneurship has evolved from being an economic term to being a dynamic way of thinking and allowing for inclusive economic growth and development (Kantis et al., 2002). It is a field that motivates individuals to venture into business opportunities despite risks (Evangelista, 2013). In support of this, Parker (2009) highlighted the positive relationship between entrepreneurship and economic growth. Hence, it is imperative for the Philippines to create an

(6)

ecosystem that will promote entrepreneurship as a lucrative occupational choice. However, for entrepreneurship to be attractive, the prospective entrepreneur must diversify market risks and increase the likelihood of venture growth (Gozun & Rivera, 2017; Rivera & Gozun, 2019).

Research Problem

As such, we probe the determinants of entrepreneurial venture growth following Storey (1994). We define entrepreneurial venture growth as the rate at which enterprises expand to the next level (Manir, 2017) in terms of workforce, customers, revenues, liquidity, profits, geographic locations, and a variety of other dimensions (Marko, 2010). Hence, we pose the research problem:

How do founder characteristics, firm attributes, and entrepreneurial strategies impact entrepreneurial venture growth? Addressing this will allow us to identify critical drivers of entrepreneurial venture growth that will aid policymakers in ensuing sustainability of entrepreneurial ventures that will aid in stimulating overall economic growth.

Research Objectives

In addressing our research problem, we set the following objectives.

1. To develop a framework that will capture determinants of entrepreneurial venture growth that will facilitate understanding of the on-going value creation by entrepreneurs.

2. To estimate the impact of founder characteristics, firm attributes, and entrepreneurial strategies on entrepreneurial venture growth.

3. To craft strategic recommendations for entrepreneurs and government on which determinant they can leverage to drive entrepreneurial venture growth.

Significance

Despite the abundance of scholarly literature on entrepreneurship, there is still little findings about entrepreneurial venture growth, especially among developing economies, like the

(7)

Philippines, due to heterogeneity of entrepreneurs and their respective ventures, as well as being in the midst of a volatile, uncertain, complex, ambiguous, and disruptive (VUCAD)-world (Rafael et al., 2020). This results in varying findings, depending on circumstance and locale. Moreover, as Gilbert et al. (2006) argued, this is an important topic because scholarly literature has focused on why ventures grow and not much on how and where growth is ensuing. Most empirical studies also explain a small portion of the variation in entrepreneurial venture growth. Hence, there is limited knowledge about the drivers of growth in most entrepreneurial ventures and cannot confidently explain growth patterns, which we would like to augment.

Our study is significant on two aspects—knowledge and policy components. For the knowledge component, it is necessary to continuously augment existing studies using alternative models and localized datasets to verify, support, and provide alternative theorization regarding entrepreneurial venture growth. For the policy component, enriching the literature on the drivers of entrepreneurial venture growth will facilitate the formulation of better policies and interventions to support entrepreneurial ventures and increase their likelihood of success.

This study is organized as follows. We conducted a literature review highlighting what has been done by scholarly studies in explaining entrepreneurial venture growth and identifying gaps that can be addressed to augment existing knowledge about this topic. We then reorganized our findings from the literature review to formulate a conceptual and empirical framework. We then operationalized our framework as applied to an appropriate dataset and derived key implications and policy recommendations to enhance entrepreneurial culture and ecosystem in the Philippines.

(8)

Literature Review Re-defining Entrepreneurial Venture Growth

The conceptualization of entrepreneurial venture growth has been predominantly associated with profitability (Knight, 1921; Kirzner, 1973), where growth is measured solely by monetary factors.

However, over time, it has been redefined to include other aspects, such as social impact and relevance (Battilana & Lee, 2014), and other non-monetary factors that are deemed essential to the overall success of a venture, which includes qualitative features such as quality of the product, market position, and relationship with customers (Hamilton, 2000). As entrepreneurial venture growth is being redefined, growth has now been largely understood as venture survival (DeSantola

& Gulati, 2017), wherein it is contextualized against the backdrop of "scaling" (Eisenmann &

Wagonfeld, 2012, p.1) or balancing internal organization and increased scope of activities that accompany growth (Chandler, 1990). More than a quantitative measure, it has been re- conceptualized to encompass other factors significant in venture success.

Factors Affecting Entrepreneurial Venture Growth

Various scholarly studies have been conducted in determining factors affecting venture growth, which were primarily hinged on the theory of venture growth or Gibrat's law. Gibrat’s law suggests that growth rates of ventures are unrelated to venture size and age; instead, it is considered a random process (Sutton, 1997). However, recent studies have shown evidence that says otherwise.

For example, Lotti et al. (2003) found evidence that Gibrat's law does not necessarily project growth patterns for small ventures. Thus, more studies continue to probe on the drivers of entrepreneurial venture growth. Storey (1994), for instance, cited three groups of determinants of entrepreneurial venture growth, namely: (a) founder characteristics, (b) firm attributes, and (c)

(9)

strategies of the entrepreneur. This section of the literature will explore the empirical determinants of venture growth.

Founder Characteristics

Certain entrepreneur's characteristics are said to influence firm growth. Schumpeter (1934) and Wickham (2006) suggested that characteristics of being creative, innovative, and being able to take risks are key to success. Although such qualities are not easily measurable, studies suggested that entrepreneur's academic background, gender, age, among other characteristics, can also be a measure of venture growth (Pajarinen et al., 2006). Parker (2009) enumerated education, experience, age, and growth motivation as variables that represent founder characteristics. On education and experience, the study of Almus and Nerlinger (1999) revealed that employment growth is higher in new ventures with founders having technology- and business-related degrees.

Korunka et al. (2011) found that founder’s gender is an important predictor of growth because it can dictate the over-all strategy of venture growth. Regarding experience, between managerial experience and previous entrepreneurial experience within the same industry, Koeller and Lechler (2006) found that managerial experience has a stronger positive impact on venture growth.

Moreover, the founder's education and work experience often serve as a source of knowledge and credibility of existing ventures, which also attract proper attention and access to information (Dencker et al., 2009). Similarly, the functional background of founders, including past company affiliations, training, and prior success, also influence the growth of ventures (Audia & Rider, 2005; Eesley et al., 2014). To an extent, a founder's personal capability is considered to be very significant in predicting venture growth (Siegel et al., 1993).

Meanwhile, Storey (1994) found mixed effects of founder’s age and previous time spent in self-employment on entrepreneurial venture growth. On growth motivation, Birley and

(10)

Westhead (1994) found no effect. It may be the case that growth aspirations should be seen in a

“negative way”—it would be surprising to hear that entrepreneurs who have no intention of growing did actually grow. In the same manner, a founder's entrepreneurial orientation can be linked to a venture's performance (Lumpkin & Dess, 1996). Following this logic, Boeker and Karichalil (2002) found that the departure of founders significantly affects a venture’s long-term performance and growth.

Firm Attributes

Papadaki et al. (2002) suggested that along with an entrepreneur's characteristics and strategies, firm characteristics also influence venture growth. These characteristics include the firm's size, age, location, and the industry and sector to which it belongs. For firm attributes, Parker (2009) enumerated variables that represent it, such as initial firm size, firm age, venture team size, limited liability, and profits. Brock and Evans (1986) and Bates (1990) found that younger firms have more variable growth rates, supporting the results of Jovanovic (1982). Cabral and Mata (2003) and Bechetti and Trovato (2002) found that funding source availability may also be a factor in venture growth where its effect is less pronounced for larger businesses. Firm attributes, such as size and age, are considered to influence venture growth where younger and smaller firms are considered to grow faster than their larger and older counterparts (Jovanovic, 1982; Evans, 1987;

Lotti et al., 2003; Calvo, 2006).

Meanwhile, Schutjens and Wever (2000) found that long-term growth tends to be higher among ventures, which commit more labor and capital resources at the time of launch. Similarly, the study of Almus and Nerlinger (1999) revealed that there is a significantly higher employment growth rate among new technology-based firms (NTBF) than non-NTBF, emphasizing that quality is the index for technology. Similarly, Nichter and Goldmark (2005) found that apart from other

(11)

factors (e.g., existing business environment), firm characteristics (e.g., technology), and finance are also considered determinants of venture growth. Venture team size is found to stimulate growth. However, just like other growth determinants, Eisenhardt and Schoonhaven (1990) argued that they are endogenous. For instance, multiple founders may have access to a broader, more heterogeneous pool of skills and experience and can provide each other with technical and psychological support, resulting in growth. Moreover, the number of founders may be positively correlated with the quality of the venture, making it endogenous. Furthermore, Variyam and Kraybill (1992) found that limited liability and multiple-establishment ventures also have higher growth rates than those who are not. However, Parker (2009) argued that these variables are endogenous. Finally, Davidsson et al. (2006) concluded that profitability and growth have ambiguous relationships. Such is the result because there is a need to distinguish between trading and retained profits (Parker, 2009). According to Watson (1990), there is a weak linkage between employment growth and trading profits. However, a strong relationship exists between employment growth and retained profits. This is because retained profits are most likely reinvested for business expansion.

Entrepreneurial Strategies

Strategies made in relation to a venture’s growth are said to have a significant effect on its future performance and growth, as suggested by the theory of path dependence (Mahoney, 2000).

The theory suggests that the decisions made on venture design can create a "lock-in" effect (Arthur 1989, p. 116). Decisions made by management teams are said to create a lasting effect on a venture's future in terms of structures, practices, and behavior (DeSantola & Gulati, 2017).

Supporting this, the theory of imprinting (Stinchcombe, 2000) suggests that founders' imprint organizations often persist over time (Johnson, 2007). Beckman and Burton (2008) found that

(12)

decision-making by the founders and the founding team can influence outcomes of ventures as enabling and constraining factors. To an extent, entrepreneurial venture growth is considered to be dependent on the strategic decisions of entrepreneurs (Davidsson, 1989), such as their choice of where and how the venture will develop (Gilbert et al., 2006). In terms of specific entrepreneurial strategies, Parker (2009) enumerated strategies for accessing multiple sources of finance, use of formal information management processes (e.g., computerization), business plans, and use of external assistance. According to the Stanford Project on Emerging Companies (SPEC) studies, strategies utilized to manage employment relations can influence organizational design in ventures (Baron et al., 1996). On the contrary, according to Shane (2003), econometric evidence is inconclusive to support the claim that business plans and planning are associated with superior venture performance. As a counterexample, the study of Bhide (1994) found that 41% of the founders in his sample had no business plan during start-up. It might be the case that the expected costs of formal planning for these entrepreneurs exceeded the expected benefits. Meanwhile, other forms of external linkages, such as franchising (Michael, 1996; Martin, 1988), connections with other enterprises (Almus & Nerlinger, 1999), and outsourcing of product distribution (Koeller &

Lechler, 2006), have shown an impact on entrepreneurial venture growth.

Others

Apart from the established factors affecting entrepreneurial venture growth, other facets also influence growth and performance. For one, government policy is considered a factor affecting venture growth where the legal environment governing the venture can become an inhibiting or supporting factor to growth (Ayegba & Omale, 2016). Similarly, cultural environment surfaces as a critical factor in influencing growth, especially in ventures located in developing economies. These cultural factors include concepts of entrepreneurship prevailing in a certain

(13)

location (Kennedy, 1976). Others point out the organizational design, which emphasizes the influence of the venture's formal structure that shapes the actions of its members and its overall operations (Galbraith, 1973; Thompson, 1967). In terms of the business environment, the level of competition in a specific industry can influence venture growth (Chamanski & Waagø, 2001).

Research Gap

Figure 1 maps out the scholarly studies we found explaining entrepreneurial venture growth. As a gap, although it is apparent that there is an abundance of studies, the state of knowledge about entrepreneurial venture growth at the individual or firm level remains to be limited, especially among developing economies (Alom et al., 2016). Moreover, despite the growing number of new ventures, the study of entrepreneurial venture growth lagged behind (DeSantola & Gulati, 2017), where the majority is focused on the role of entrepreneurial venture growth to organizational change (Weber, 1946; Blau et al., 1966; Kimberly & Miles, 1980). A plausible reason for this is the evident heterogeneity of entrepreneurs and their ventures; hence, inconclusive findings. It is also apparent that empirical studies can just explain a portion of the variation in entrepreneurial venture growth. As such, we still know little about the drivers of growth in most entrepreneurial ventures and cannot confidently predict growth patterns. Hence, it is imperative to continuously augment existing studies using alternative models and localized datasets. Enriching the literature on the drivers of entrepreneurial growth ventures will allow us to formulate better policy recommendations to promote new entrepreneurial ventures.

(14)

Figure 1 Literature Map

Source: Constructed by the authors

(15)

Framework and Methodology Conceptual Framework

In addressing our first research objective of developing a framework that will capture determinants of entrepreneurial venture growth, we appeal to the framework of Shah et al. (2013), as seen in Figure 2.

Figure 2

Factors Affecting Entrepreneurial Growth

Source: Shah et al. (2013)

Given our problem statement, we modify the framework of Shah et al. (2013), as seen in Figure 3. Instead of internal and external factors driving entrepreneurial growth, we adapt the determinants enumerated by Storey (1994). These comprise both internal and external factors.

Importantly, most constructs are represented by variables captured by our chosen dataset – the Global Entrepreneurship Monitor (GEM) Adult Population Survey (APS) data for the Philippines.

We summarized the bases of our framework in Table 1 by indicating a priori expectations.

These will be established and verified through the empirical framework.

(16)

Figure 3

Factors Affecting Entrepreneurial Venture Growth

Source: Constructed by the authors

Entrepreneurial Venture Growth (e.g., growth in workforce, market,

financial position and performance) Founder

Characteristics (e.g., education, experience, age, gender, growth

motivation)

Firm Attributes (e.g., firm size, firm age, venture team size, limited liability, profits)

Entreprenurial Strategies (e.g., business plans,

external assistance, market expansion, external participation)

(17)

Table 1

A Priori Expectations on Entrepreneurial Venture Growth Construct/Variable A-priori

Expectations Source

Founder Characteristics (Storey, 1994; Parker 2009) Education

EDCFNi + Siegel et al. (1993); Almus and Nerlinger (1999);

Dencker et al. (2009).

Experience OCCFNi

+/- Almus and Nerlinger (1999); Koeller and Lechler (2006)

+ Siegel et al. (1993); Audia and Rider (2005);

Dencker et al. (2009); Eesley et al. (2014) Age

AGEFNi +/- Storey (1994)

Gender GDRFNi

+/- Korunka et al. (2011)

Growth Motivation MTVFNi

+ Lumpkin and Dess (1996)

0 Birley and Westhead (1994)

Firm Attributes

(Storey, 1994; Parker, 2009; Papadaki et al., 2002) Initial Firm Size

NMOWNi - Jovanovic (1982); Evans (1987); Lotti et al.

(2003); Calvo (2006) Firm Age

FRAGEi -

Jovanovic (1982); Brock and Evans (1986); Evans (1987); Bates (1990); Lotti et al. (2003); Calvo

(2006) Venture Team Size

TMSZEi - Jovanovic (1982); Evans (1987); Lotti et al.

(2003); Calvo (2006) Entrepreneurial Strategies

(Storey, 1994; Mahoney, 2000; DeSantola & Gulati, 2017) Business Plans

MKEXPi

TECHNi

COMPTi

0 Shane (2003)

+ Bhide (1994); Kumaran (2018)

+/-

Davidsson (1989); Baron et al. (1996); Nichter and Goldmark (2005); Beckman and Burton

(2008) External

Participation INTORi

EXPORi

+ Michael (1996), Martin (1988), Almus and Nerlinger (1999), Koeller and Lechler (2006)

(18)

Empirical Framework

We transformed our framework illustrated in Figure 2 and Table 1 into an economic model as seen in Equation 1. Classifications made by Storey (1994) served as the basis of our multivariate econometric model, with results drawn from reduced-form growth models. However, according to Wiklund (2007), reduced-form models suffer from endogeneity problems. That is, some cross- sectional characteristics (i.e., individual peculiarities, firm attributes) might indirectly influence growth through stochastic entrepreneurial strategic choices.

EVGi = f(vFOCHRi, vFRATTi, vESMEXi, vESEPNi) (1)

where EVGi is entrepreneurial venture growth, vFOCHRi is a vector containing founder characteristics, vFRATTi is a vector containing firm attributes, vESMEXi is a vector containing entrepreneurial strategies on market expansion, and vESEPNi is a vector containing entrepreneurial strategies on external participation. From these constructs, we matched them to the variables available in GEM. Hence, we are able to come up with the elements of each vector as shown in Equations 2, 3, 4, and 5, respectively. All are specifically captured and measured by GEM.

vFOCHRi = [education, experience, age, gender, motivation] (2) vFRATTi = [firm size, team size, firm age] (3) vESMEXi = [market expansion, technological level, competition] (4) vESEPNi = [international orientation, export orientation] (5)

We rewrite Equation 1 into its econometric form as seen in Equation 6, where ui is the stochastic disturbance term capturing all other possible growth drivers not captured by our model.

EVGi = f(vFOCHRi, vFRATTi, vESMEXi, vESEPNi) + ui (6)

Table 2 shows the corresponding variables measuring the constructs enumerated in Equations 1 to 6.

(19)

Table 2

Constructs and Variables, as Measured by GEM

Equation 6 Construct Definition Representation in GEM data

Variable in

our model Description Categories

included Metric Endogenous

variable EVGi

Entrepreneurial Venture Growth

Rate at which enterprises expand

to the next level (Manir, 2017). TEAJOBGR EJG5Yi

Growth assessed in terms of employees, customers, revenue, liquidity, profit, geographic locations and a variety of other

dimensions (Marko, 2010).

Discrete

Expected job growth in persons

in 5 years

Founder characteristics

vFOCHRi

Education

Represents the technical skills of an entrepreneur; for demographic

profiling.

GEMEDUC EDCFNi

Dummy variable indicating highest educational attainment by entrepreneurial

venture’s founder

UPSSEi

Up to some secondary

education SECDEi Secondary degree POSECi Post-secondary GRADXi Graduate

experience

Occupation

Represents the industry experience of the entrepreneur;

for demographic profiling.

GEMOCCU OCCFNi Dummy variable indicating work status of the entrepreneurial venture’s founder

FULLTi

Full or part time work (including self-employment) PARTTi Part time work

only RETDSi Retired / disabled HOMEMi Homemaker

STDNTi Student NOTWKi Not working

OTHRSi Others UKOCCi Unknown

occupation

Age

Represents the length of experience of the entrepreneur;

for demographic profiling.

AGE AGEFNi

Age in years of the entrepreneurial venture’s founder (a squared term, AGEFN2i, is added in the regression to

allow us to model the non-linear relationship between age and any independent variable (Gujarati & Porter,

2009).

Discrete

Entrepreneurial venture’s founder’s

age in years

Gender For demographic profiling. GENDER GDRFNi GMALEi Male

(20)

Dummy variable indicating whether entrepreneurial venture’s founder is either

male or female

GFMALi Female

Motivation

Represents the driving factor why entrepreneurial venture’s founder

engaged in doing business.

TEAyyMOT (where yy indicates year of

survey period)

MTVFNi

Dummy variable indicating whether the motivation is opportunity (increase income,

financial independence) or necessity (maintain income for sustained

consumption)

MTOPPi Opportunity motive MTNECi Necessity motive

UKMTVi Unknown motive

Firm attributes vFRATTi

Firm size

Firm size has also been used to proxy for numerous theoretical constructs ranging from risk to liquidity or even political costs

(Ball & Foster, 1982).

TEAOWNER NMOWNi

Usually measured by turnover, balance sheet accounts, and number of employees to indicate whether a firm is micro, small, or medium. However, firm size remains a poorly defined concept Trigueiros (2000).

Empirical studies typically revert to proxies such as number of employees, total

assets, sales, or market capitalization.

Discrete Number of owners in the firm

IPTEAMSIZE TMSZEi Discrete

Number of members in the

team

Firm age Represents how long an enterprise has been existing.

BABYBUSO

FRAGEi

Dummy variable indicating what kind of enterprise is based on how long it has been

managed and operated

UP42Mi

Baby business (manages and owns

a business that is up to 42 months

old),

ESTBBUSO OL42Mi

Established business (manages

and owns a business that is

older than 42 months).

UKFRAi Unknown

Entrepreneurial strategy: Market

expansion vESMEXi

Market expansion

A strategy is a consciously intended course of action to deal

with a situation (Mintzberg, 1987). Market expansion activities include methods such as

planning and marshaling resources for their most efficient and effective use to bring about a

TEAyyMEM (where yy indicates year of

survey period)

MKEXPi

Dummy variable indicating market expansion strategy implemented by the

enterprise

NOMKXi

No market expansion

SMXNTi

Some market expansion, no new

technology

SMXWTi

Some market expansion, with new technology

(21)

desired future, such as achievement of a goal or solution

to a problem.

PFMKXi Profound market expansion UKMXPi Unknown market

expansion

Technology

level TEAHITEC TECHNi Dummy variable indicating technology

level employed by the enterprise

LOTECi No/low technology MDTECi Medium

technology HITECi High technology UKTECi

Unknown technology

Competition SUCOMPET COMPTi Dummy variable indicating degree of

competition

MNYBCi Many business competitors FEWBCi Few business

competitors NONBCi No business competitor UKCOMi Unknown

competition

Entrepreneurial strategy:

External participation

vESEPNi

International orientation

Degree of exposure or participation in the foreign

market.

TEAEXP4C INTORi

Dummy variable indicating degree of international orientation measured by the

share of foreign customers in output

>=76Pi 76% to 100%

2675Pi 26% to 75%

0125Pi 1% to 25%

ZEROPi none

UKINTi

Unknown international

orientation

Export

orientation SUEXPORT EXPORi

Dummy variable indicating degree of export orientation measured by the

percentage of output for exports

MO90Pi More than 90%

7690Pi 76% to 90%

5175Pi 51% to 75%

2650Pi 26% to 50%

1125Pi 11% to 25%

LE10Pi 10% or less ZERPRi None UKEXPi Unknown export

orientation

(22)

From Table 2, we can rewrite Equations 1 to 6 into Equations 7 to 12, respectively, to indicate all variables measuring our constructs.

EJG5Yi = f(vFOCHRi, vFRATTi, vESMEXi, vESEPNi) (7) vFOCHRi = [EDCFNi, OCCFNi, AGEFNi,AGEFN2i, GDRFNi,MTVFNi,] (8) vFRATTi = [NMOWNi,TMSZEi, FRAGEi] (9) vESMEXi = [MKEXPi,TECHNi,COMPTi] (10)

vESEPNi = [INTORi,EXPORi] (11) EJG5Yi = f(EDCFNi, OCCFNi, AGEFNi, AGEFN2i,GDRFNi,MTVFNi,

NMOWNi,TMSZEi, FRAGEi,MKEXPi,TECHNi,COMPTi,INTORi,EXPORi) + ui (12)

Dataset

In addressing our second research objective of estimating the impact of founder characteristics, firm attributes, and entrepreneurial strategies on entrepreneurial venture growth, we would be subjecting the GEM APS for the Philippines, covering years 2006, 2013, 2014, and 2015, to Equation 12. Because entrepreneurs are not alike, the GEM APS is a unique instrument administered by GEM National Teams to a representative national sample of at least 2,000 respondents. The following were examined: (a) characteristics, motivations, and ambitions of individuals starting businesses; (b) level and nature of entrepreneurial activities around the world;

and (c) social attitudes towards entrepreneurship (https://www.gemconsortium.org/data).

Alternatively, it explores the role of the individual in the life cycle of the entrepreneurial process by probing on business characteristics, people’s motivation for starting a business, actions taken to start and run a business, and entrepreneurship-related attitudes (https://www.gemconsortium.org/wiki/1141).

Following Gozun and Rivera (2016, 2017) and Rivera and Gozun (2019), GEM is an appropriate dataset for our purposes because it takes a wide-ranging perspective of what it acknowledges as business activity. It does not discriminate between old and newly established and registered business because it subscribes to the occupational dimension of entrepreneurship.

(23)

Moreover, it also looks into entrepreneurship from a behavioral perspective by identifying employees within organizations who behave entrepreneurially, either intrapreneurship or corporate entrepreneurship. Furthermore, it also captures the combination of the stages of nascent entrepreneurship and owning-managing a new firm or early-stage entrepreneurial activity.

Methodology

Given the specification of Equation 12, we recognize the possibility of the endogeneity problem. It exists when a parameter or variable and the error term are correlated. According to Gujarati and Porter (2009), this may happen due to errors in measurement, autoregression with autocorrelated errors, simultaneity bias, sample selection errors, and omitted variables. Moreover, this is also likely when cross-sectional characteristics indirectly influence growth through stochastic entrepreneurial strategic choices (Wiklund, 2007).

Heteroscedasticity (i.e., inconstant variance) is also likely to be present because the GEM is cross-sectional data. According to Gujarati and Porter (2009), although this does not cause ordinary least squares (OLS) coefficient estimates to be biased, it can the variance of the estimated OLS coefficients to underestimate or overestimate the population variance. That is, regression analysis using heteroscedastic data can still generate an unbiased estimate for the relationship between exogenous and endogenous variables. However, standard errors and inferences derived would be spurious. Hence, biased standard errors makes inferential statistics unreliable.

Therefore, we would utilize the linear generalized method of moments (GMM) estimation technique to analyze the impact of founder characteristics, firm attributes, and entrepreneurial strategies on entrepreneurial venture growth. According to Baum et al. (2003), when faced with heteroscedasticity of unknown form, the GMM introduced by Hansen (1982) is advised. It employs orthogonality conditions to allow for efficient estimation given an unknown form of

(24)

heteroscedasticity. Also, many standard estimators, including the instrumental variable (IV) and OLS, are deemed as subsets of GMM estimators. Thus, with heteroscedasticity, the GMM estimator is more efficient than any other estimator (Baum et al., 2003).

Our preference towards the GMM estimation technique is also due to its robustness to differences in data generating process (DGP) specifications. It also automatically addresses endogeneity. Under the GMM, a sample mean or variance estimates its population counterpart regardless of the underlying process (Greene, 2003). Thus, it provides flexibility from unnecessary distributional assumptions (e.g., normality assumption under OLS). However, we underscore that this has a cost. If more is known about the DGP, such as its specific distribution, then the GMM may not utilize all available information. Consequently, the estimates become inefficient. Hence, according to Greene (2003), the maximum likelihood estimation (MLE) is deemed the alternative approach because it utilizes out-of-sample information and provides more efficient estimates.

Likewise, although we recognize that our empirical specification is quite cumbersome due to the number of categorical variables included, this may result in the model being not identified, initial weight matrix being not positive semi-definite, or iterations are non-convergent due to non- concavity. To address this, we would be regressing our endogenous variable against each of our vectors separately. Moreover, should any of the issues arise resulting in the failure of GMM, we would then resort to OLS and its accompanying post-regression tests for violations against the classical linear regression model (CLRM). Both GMM and OLS estimator will have the same coefficient because GMM is a class of estimators that include OLS (Greene, 2003). That is, a GMM estimator can be constructed that is equivalent to the OLS estimator.

(25)

Results and Discussion

Prior to the presentation and discussion of our regression results, we present in Table 3 descriptive statistics for discrete variables and cross-tabulations for nominal variables in Table 4.

These will provide a glimpse of our data’s peculiarities, which will aid in understanding the statistical and practical implications of our regression results.

Descriptive Statistics

We interpret our descriptive statistics generally across all survey periods unless a peculiarity is observed for a specific year. Table 3 shows that survey periods 2006, 2014, and 2015 have 2,000 respondents, whereas 2013 has 2,500 respondents. Although this is the case, only approximately 20% have been used in data analysis. This is because not all respondents have an answer for EJG5Yi. This is also why TMSZEi has not been very useful in data analysis because of data unavailability (for 2006) and data insufficiency (for 2013, 2014, and 2015). We also note that respondents for all survey periods are those members of the labor force population, aged 18 to 64, with the majority from the 21–30 and 31–40 brackets. This is indicative of the entrepreneurship of the youth (Gozun & Rivera, 2017; Rivera & Gozun, 2019). Of course, there is also a significant number of entrepreneurs at much higher age brackets, which is indicative of the entrepreneurial activities of those who have the experience, financial stability, and extensive networks (St. Pierre, 2017; Gaskell, 2019).

Table 3

Cross Tabulations and Descriptive Statistics

2006 N Mean Standard Deviation Minimum Maximum

EJG5Yi 426 1.7441 5.1585 -4 80

<0 4 -2.0000 1.4142 -4 -1

0 222 0.0000 0.0000 0 0

1 to 10 189 2.5026 1.7674 1 10

11 to 20 7 15.8571 2.6726 12 20

>20 4 41.7500 26.0560 22 80

2006 N Mean Standard Deviation Minimum Maximum

(26)

AGEFNi 2,000 37.8515 12.2732 18 64

<21 137 19.0073 0.8617 18 20

21-30 524 25.7748 2.8937 21 30

31-40 540 35.6722 2.8782 31 40

41-50 449 45.2428 2.8957 41 50

51-60 274 55.6350 2.8680 51 60

>60 76 62.7895 1.0239 61 64

2006 N Mean Standard Deviation Minimum Maximum

NMOWNi 426 1.3404 0.8369 1 10

1 325 1.0000 0.0000 1 1

2 78 2.0000 0.0000 2 2

3 16 3.0000 0.0000 3 3

>3 7 6.0000 2.2361 4 10

2006 N Mean Standard Deviation Minimum Maximum

TMSZEi . . . . .

1 . . . . .

2 . . . . .

3-10 . . . . .

>10 . . . . .

2013 N Mean Standard Deviation Minimum Maximum

EJG5Yi 474 2.3439 15.0571 -9 300

<0 15 -3.0000 3.0237 -9 -1

0 248 0.0000 0.0000 0 0

1 to 10 202 2.7574 2.3278 1 10

11 to 20 4 18.0000 4.0000 12 20

>20 5 105.4000 111.8740 30 300

2013 N Mean Standard Deviation Minimum Maximum

AGEFNi 2,500 37.3752 12.7289 18 64

<21 207 18.9903 0.8418 18 20

21-30 674 25.4867 2.7513 21 30

31-40 660 35.3333 2.8670 31 40

41-50 480 45.1792 2.8704 41 50

51-60 359 55.1532 2.6744 51 60

>60 120 62.6917 1.1212 61 64

2013 N Mean Standard Deviation Minimum Maximum

NMOWNi 474 1.4958 1.2546 1 10

1 337 1.0000 0.0000 1 1

2 102 2.0000 0.0000 2 2

3 18 3.0000 0.0000 3 3

>3 17 6.7059 2.8889 4 10

2013 N Mean Standard Deviation Minimum Maximum

TMSZEi 23 4.1739 4.1522 1 20

1 5 1.0000 0.0000 1 1

2 4 2.0000 0.0000 2 2

3-10 13 4.8462 2.1153 3 10

(27)

>10 1 20.0000 . 20 20

2014 N Mean Standard Deviation Minimum Maximum

EJG5Yi 382 4.2304 46.3254 -2 900

<0 12 -1.4167 0.5149 -2 -1

0 199 0.0000 0.0000 0 0

1 to 10 160 2.8500 2.1782 1 10

11 to 20 6 15.5000 2.4290 12 19

>20 5 216.8000 382.7254 24 900

2014 N Mean Standard Deviation Minimum Maximum

AGEFNi 2,000 36.8260 12.8131 18 64

<21 191 18.8901 0.8482 18 20

21-30 565 25.3788 2.8821 21 30

31-40 475 35.2653 2.8546 31 40

41-50 410 45.1463 2.7634 41 50

51-60 265 54.9472 2.6452 51 60

>60 94 62.5851 1.1303 61 64

2014 N Mean Standard Deviation Minimum Maximum

NMOWNi 382 1.6021 1.3414 1 10

1 259 1.0000 0.0000 1 1

2 79 2.0000 0.0000 2 2

3 25 3.0000 0.0000 3 3

>3 19 6.3158 2.4507 4 10

2014 N Mean Standard Deviation Minimum Maximum

TMSZEi 41 15.7073 23.3968 1 100

1 5 1.0000 0.0000 1 1

2 6 2.0000 0.0000 2 2

3-10 18 5.3889 2.1182 3 10

>10 12 44.1667 26.9270 15 100

2015 N Mean Standard Deviation Minimum Maximum

EJG5Yi 394 13.8452 144.6942 -5 1,998

<0 13 -1.7692 1.3634 -5 -1

0 164 0.0000 0.0000 0 0

1 to 10 201 2.8806 2.2728 1 10

11 to 20 8 13.1250 1.6421 11 15

>20 8 599.2500 880.4680 30 1,998

2015 N Mean Standard Deviation Minimum Maximum

AGEFNi 2,000 37.9095 12.5787 18 64

<21 151 18.8808 0.8240 18 20

21-30 501 25.3952 2.7327 21 30

31-40 533 35.3771 2.8906 31 40

41-50 429 45.3800 2.8885 41 50

51-60 299 55.0836 2.8230 51 60

>60 87 62.6552 1.0979 61 64

(28)

NMOWNi 394 1.5330 1.2149 1 10

1 267 1.0000 0.0000 1 1

2 97 2.0000 0.0000 2 2

3 13 3.0000 0.0000 3 3

>3 17 6.1176 2.4719 4 10

2015 N Mean Standard Deviation Minimum Maximum

TMSZEi 77 5.0519 7.6397 0 60

1 12 1.0000 0.0000 1 1

2 18 2.0000 0.0000 2 2

3-10 40 5.1500 2.1549 3 10

>10 5 27.0000 18.9077 15 60

Ultimately, throughout the various survey periods, we have also seen that respondents have indicated contraction, status quo, or expansion in their job growth. This is indicative of the varying growth prospects of business ventures with respect to employment creation. However, it is apparent that a large proportion of entrepreneurs expected to create more jobs than they can actually do. We found entrepreneurs who expected job growth to be as much as by the hundreds (for 2006, 2013, and 2014) to thousands (for 2015). According to Shane (2012), we should exercise caution in interpreting such result because this “is much higher than the share of entrepreneurs that actually has a high growth company” (para. 3)4 and “overstates the share of new businesses that are ‘high growth’” (para. 4)5. Hence, “if only about 1 out of every 20 entrepreneurs who expect to employ 20 or more people when their businesses are five years old actually does so”, then entrepreneurs tend to overestimate their job creation capabilities as well as their business’ survival, sales and profits of their businesses (Shane, 2012, para. 5).

4 According to Shane (2012), the Census’ Business Dynamics database indicated that only 2% of five-year-old companies have 20 or more employees.

5 Shane (2012) furthered that the Census’ Business Dynamics database also revealed that slightly less than half of new businesses survive to age five. That is, “adjusting the share of surviving five-year-old businesses with 20 or more employees by the failure rate of new companies reveals that less than 1% of businesses started in a given year have 20 or more employees at the time of their fifth birthday” (par. 4).

(29)

Hence, our initial findings from the descriptive statistics warranted us to exercise caution and address this overestimation by discounting entrepreneurs’ job growth projections.

Cross Tabulations

Because our data is from a survey, as part of the analysis, it is appropriate to do cross- tabulations or contingency tables (Lind et al., 2006). These are tables that present the results of the entire group and results from sub-groups of survey respondents. According to De Franzo (n.d.), this will enable us to examine relationships within the data that might not be readily apparent when analyzing total survey responses. We tabulated in Table 4 all our variables of interest against EJG5Yi. We interpret our cross-tabulations generally across all survey periods unless a peculiarity is observed for a specific year.

From Table 4, with respect to founder characteristics, we can see that most of the entrepreneurs in our sample have secondary and post-secondary education, which is indicative of the degree of technical skills Filipino entrepreneurs have. In fact, according to Lavinsky (2014), entrepreneurs are educated, contradicting the “growing misconception that higher education is not needed for – and may even inhibit – entrepreneurial success” (Arruda, 2018, para. 1). Although, there are generally more female entrepreneurs than male. This supports the discussion of Castrillon (2019) that women are turning to entrepreneurship for the following reasons: to have more flexibility, to charge what they are worth, to have more control over their future, to advance more quickly, and to follow their passion. From our discussion, it may follow that a significant proportion have full-time employment, which is also indicative of entrepreneurship being used either as passive income or multiple revenue streams (Constable, 2018). Consistently, our distribution also reveals that Filipino entrepreneurs engage in a business due to their opportunistic motive, more than the necessity motive. According to Juneja (n.d.), entrepreneurial success is

(30)

driven by the entrepreneur’s ability to create opportunities and be opportunistic – being “in control over their future wherein they were able to sense the future as well as leverage on the intersecting processes of computing, technological change, and changing workplace processes” (para. 14).

We also emphasize that as per our distribution across the vector of founder characteristics, Filipino entrepreneurs do have a modest estimate of their venture’s job growth (i.e., within 0 to 10). This is not reflective of the overoptimism and overestimation highlighted by Shane (2012).

With respect to firm attributes, we can see from Table 4 that most of the entrepreneurial ventures are either sole proprietorship or partnership whose businesses are classified as “baby business,” which by GEM definition means a business that is managed and owned up to 42 months.

Such observations reflect the state and composition of MSMEs in the Philippines, as per DTI (2020). Regardless of ownership and firm age, Filipino entrepreneurs still have a modest estimate of their venture’s job growth, which does not support the arguments of Shane (2012).

With respect to entrepreneurial strategy (market expansion), we can see from Table 4 that most Filipino entrepreneurs have zero to some market expansion, employing low technology.

Likewise, most are implementing a red ocean strategy (i.e., many business competitors), and there also many who are not aware of the kind of competition they are in. This reflects the need for MSMEs in the country to be uplifted in terms of DTI’s 7Ms of successful entrepreneurs6 (University of the Philippines Institute for Small-Scale Industries – Diliman, 2020). Market expansion accompanied by technology can solidify the growth trajectory of an entrepreneurial venture because it can achieve more profits with less investments and allow for better communication, internationalization, and networking (Kumaran, 2018). It is also important to note

6 DTI’s 7Ms of successful entrepreneurs are: (1) mindset change, (2) mastery, (3) mentoring, (4) money, (5) machine, (6) market access, and (7) models of business. See https://www.dti.gov.ph/negosyo/the-7ms-of-successful- entrepreneurs/ for the full details.

(31)

that regardless of strategy, technology, and competition, Filipino entrepreneurs still have a modest estimate of their venture’s job growth, which contradicts the arguments of Shane (2012).

Finally, with respect to entrepreneurial strategy (external participation), we can see from Table 4 that most Filipino entrepreneurs have zero to a little international and export orientations.

This is indicative of the size and productive capacities of Philippine-MSMEs, which can cater mostly, if not fully, to domestic demand because they are not connected to global value chains (Francisco et al., 2018). Connecting Philippine-MSMEs to global value chains has its own obstacles and challenges, which need to be hurdled to enhance external participation that can facilitate entrepreneurial venture growth. Similar to our earlier observations, because of low international and export orientation, there is not much job growth expected.

Our cross-tabulations also serve as contingency tables, allowing for the implementation of a contingency table analysis (i.e., chi-square test of independence). This is done “to formally test for a relationship between two nominal-scaled variables” (Lind et al., 2006, p. 476) where the null hypothesis is independence (i.e., no relationship between the two nominal-scaled variables of interest). However, we would not implement this anymore as we will proceed immediately with the regression, which can capture the information we can derive from this.

(32)

Table 4

Cross Tabulations

Year 2006 2013

EJG5Yi <0 0 1 to 10 11 to 20 >20 Total <0 0 1 to 10 11 to 20 >20 Total

vFOCHRi

EDFCNi

UPSSEi 2 81 51 2 0 136 0 0 0 0 0 0

SECDEi 0 59 58 0 2 119 10 120 94 2 1 227

POSECi 2 81 80 5 2 170 3 72 52 2 1 130

GRADXi 0 1 0 0 0 1 0 1 1 0 0 2

UKEDUi 0 0 0 0 0 0 2 55 55 0 3 115

Total 4 222 189 7 4 426 15 248 202 4 5 474

OCCFNi

FULLTi 3 165 146 7 2 323 15 229 175 3 4 426

PARTTi 1 57 43 0 2 103 0 3 4 0 0 7

RETDSi 0 0 0 0 0 0 0 1 0 0 0 1

HOMEMi 0 0 0 0 0 0 0 3 11 0 0 14

STDNTi 0 0 0 0 0 0 0 1 1 0 0 2

NOTWKi 0 0 0 0 0 0 0 5 3 0 0 8

OTHRSi 0 0 0 0 0 0 0 0 0 0 0 0

UKOCCi 0 0 0 0 0 0 0 6 8 1 1 16

Total 4 222 189 7 4 426 15 248 202 4 5 474

AGEFNi

<21 0 3 4 0 0 7 0 8 10 0 0 18

21-30 0 61 42 3 1 107 2 51 54 1 1 109

31-40 2 64 63 1 3 133 4 75 56 1 2 138

41-50 2 52 52 2 0 108 5 57 40 1 0 103

51-60 0 33 21 1 0 55 3 44 32 1 0 80

>60 0 9 7 0 0 16 1 13 10 0 2 26

(33)

Total 4 222 189 7 4 426 15 248 202 4 5 474 GDRFNi

GMALEi 1 96 85 4 3 189 7 89 71 1 2 170

GFMALi 3 126 104 3 1 237 8 159 131 3 3 304

Total 4 222 189 7 4 426 15 248 202 4 5 474

MTVFNi

MTOPPi 3 108 109 5 4 229 9 141 117 2 4 273

MTNECi 1 114 80 2 0 197 6 105 83 2 1 197

UKMTVi 0 0 0 0 0 0 0 2 2 0 0 4

Total 4 222 189 7 4 426 15 248 202 4 5 474

vFRATTi

NMOWNi

1 3 177 137 5 3 325 11 192 131 2 1 337

2 0 33 43 2 0 78 3 42 54 2 1 102

3 1 8 7 0 0 16 1 9 7 0 1 18

>3 0 4 2 0 1 7 0 5 10 0 2 17

Total 4 222 189 7 4 426 15 248 202 4 5 474

TMSZEi

1 . . . 0 0 1 0 0 0 1

2 . . . 0 0 0 1 0 0 1

3 to 10 . . . 0 0 0 3 0 1 4

>10 . . . 0 0 0 0 1 0 1

Total 0 0 0 0 0 0 0 1 4 1 1 7

FRAGEi

UP42Mi 2 161 147 7 3 320 13 117 59 2 4 195

OL42Mi 0 9 10 0 0 19 0 0 4 0 0 4

UKFRAi 2 52 32 0 1 87 2 131 139 2 1 275

Total 4 222 189 7 4 426 15 248 202 4 5 474

vESMEXi MKEXPi

NOMKXi 3 160 111 2 2 278 8 98 65 0 0 171

SMXNTi 0 24 31 2 0 57 3 69 59 0 3 134

SMXWT 1 38 44 3 1 87 3 59 66 4 2 134

Pigura

Figure 1   Literature Map
Table 6. Regression details

Mga Sanggunian

NAUUGNAY NA DOKUMENTO

Moreover, a critical important aspect in evaluating the performance of the tests in terms of their power is the change of the parameter values for ϕ11 as for the parameter value for ϕ21

In furtherance of our disclosure last 24 January 2022, be advised that, today, the Far Eastern University “FEU”, Jerudong Park Medical Centre “JPMC” Sendirian Berhad, and JPMC College