International Review of Financial Analysis 75 (2021) 101722

Contents lists available at ScienceDirect

International Review of Financial Analysis

journal homepage: www.elsevier.com/locate/irfa

The role of trust and social commitment in start-up financing

Viviana Fernandez *
Business School, Universidad Adolfo Ibañez, Av. Diagonal Las Torres 2700, Office 512–C, Santiago, Chile

A R T I C L E I N F O A B S T R A C T

Editor: Prof. Brian M. Lucey Access to finance is a perennial problem for business start-ups. This article advances the extant literature on the
determinants of entrepreneurial finance by presenting a comprehensive analysis of the amount needed to start a
JEL classification: business, the amount of self-investment and the number of external sources of financing utilized. Based on in­
L26 formation from the Global Entrepreneurship Monitor 2015, the empirical analysis shows that the availability of
M13
external sources of financing depend primarily on demographics, entrepreneurial experience, business features,
Keywords: e.g., export propensity and growth potential; and, on the phase of economic development. Furthermore, rela­
Entrepreneurial finance
tively high levels of social trust may increase the likelihood of obtaining external financing. The novelty of the
Social trust
product/service offered, financial barriers and investor/property rights protection in turn may also explain the
Social entrepreneurship
Gender proportion of self-investment provided to a new business. An analysis of financing options of socially-oriented
Global entrepreneurship monitor businesses shows that they are more successful in obtaining financial support from private investors/venture
capitalists and governments.

1. Introduction
In business studies, entrepreneurship is customarily associated with
opportunity recognition and new venture creation (Parker, 2018, page
7). A key aspect to a new endeavor is its financing. In this regard,
traditional models of entrepreneurial finance include informal invest­
ment through the founders themselves (i.e., self-financing), borrowing
from friends, family and colleagues, public funding (seed capital) and
venture capital investments in promising entrepreneurial firms, partic­
ularly in developed countries. Over the past decade, newer financing
models—including business angels, microfinance, and small business
accelerators, have matured considerably, while new financing vehicles,
such as peer-to-peer lending and crowd-funding have emerged as pop­
ular alternatives (e.g., Bellavitis, Filatotchev, Kamuriwo, & Vanacker,
2017; Cumming, Deloof, Manigart, & Wright, 2019; Cumming & Groh,
2018; Cumming & Johan, 2017; Daniels, Herrington, & Kew, 2016; Di
Pietro & Buttice, 2020; Mollick, 2014; Oranburg, 2020).
In particular, entrepreneurial finance differs from traditional
corporate finance in two major aspects (e.g. Cumming et al., 2019).
First, obtaining external financing is difficult because of information
asymmetries between entrepreneurs and finance providers—e.g., diffi­
culty for investors to judge the business quality; outside investors and
entrepreneur may disagree on the value of the project and its likelihood
of success. In addition, finance providers face moral hazard problems
relating to their investment because once external funds are obtained,
entrepreneurs may misuse or misallocate them for their personal benefit.
Such moral hazard issues get exacerbated when investor protection is
low. Second, entrepreneurial firms often lack tangible and intangible
assets, such as human capital and access to networks, reputation or
legitimacy, necessary to achieve value creation potential.
Usual solutions to overcome these issues, such as investor moni­
toring, allocation of contractual rights, capital staging, and risk-sharing,
may be inadequate to overcome information asymmetries in early-stage
start-ups. Instead, information transfer through social ties and social
obligation may prove suitable. Indeed, ongoing social relationships help
to solve agency problems through both parties’ commitment to fairness
and trust, particularly when legal protection is weak (e.g., Denis, 2004;
Ding, Au, & Chiang, 2015; Nofsinger & Wang, 2011; Pruthi, 2014). In
this regard, Mickiewicz and Rebmann (2020) distinguish between
particular trust and extended trust. The former can arise either through
interactions or can be based on membership of a certain group. Extended
trust in turn involves generalized trust or trusting the general population
due to widely shared norms that discourage opportunism; and, institu­
tional trust grounded on the existence of formal mechanisms by which
transactions are protected and regulated.
Besides social ties, information asymmetries can be ameliorated by
signals sent by the entrepreneur about his/her venture through product
type and production technology (new versus existing), and

* Corresponding author.
E-mail address: viviana.fernandez@uai.cl.

https://doi.org/10.1016/j.irfa.2021.101722
Received 26 January 2021; Received in revised form 24 February 2021; Accepted 25 February 2021
Available online 9 March 2021
1057-5219/© 2021 Elsevier Inc. All rights reserved.
V. Fernandez
entrepreneurial skills/experience. Such characteristics represent valu­
able information to assess the value creation potential of initial start-
ups, particularly in low information environments (Nofsinger & Wang,
2011)
The core of this study relies on information collected by the Global
Entrepreneurship Monitor (GEM) Adult Population Survey (APS) in
2015 on financing models used by early-stage/established businesses in
62 countries around the world. GEM is a networked consortium of na­
tional country teams primarily associated with leading academic in­
stitutions, which collects data on entrepreneurship directly from
individual entrepreneurs (https://www.gemconsortium.org/about
/gem/5). GEM encompasses early-stage entrepreneurship in the Total
Early-Stage Entrepreneurial Activity (TEA) rate, which includes nascent
and new entrepreneurial activities. Specifically, TEA measures the
prevalence of individuals engaged in nascent entrepreneurship and new
enterprise ownership in the adult population (18–64 years old). Nascent
entrepreneurs are those who have taken steps to start a new business,
but have not yet paid salaries or wages for more than three months. New
entrepreneurs in turn are those running a business that has been in
operation for between 3 and 42 months (3.5 years).
The contribution of this study is two-fold. First, it advances the
extant literature on entrepreneurial financing (e.g., Brown & Rocha,
2020; Frešer & Tominc, 2018; Lee, Sameen, & Cowling, 2015; Nofsinger
& Wang, 2011; Roper & Scott, 2009; Vaznyte & Andries, 2019) by
modeling the determinants of alternative sources of financing (e.g.,
personal savings, family and friends, banks, government) utilized by
early-stage entrepreneurs located in a wide sample of countries. To that
end, new statistical approaches and a more comprehensive analysis is
offered by incorporating the role of social trust (e.g., Ding et al., 2015) in
obtaining external financing. In our view, offering international evi­
dence is relevant because it allows a richer perspective of access to and
drivers of start-up external financing, as opposed to the evidence from a
single country (e.g., the United States, China, and Germany) which is
predominant in the existing literature.
In regards to access to financing, OECD figures (https://stats.oecd.
org/) on venture capital investment show that member economies
spent an average of 3.2% and 2.4% of GDP on start-up/other early-stage
and later-stages ventures, respectively, during 2007–2019. However,
there was substantial heterogeneity across members during this period.
For example, while Canada, Finland, the United States, and Sweden
spent an average of 5.3%, 3.4%, 12.4%, and 3.7% of GDP on start-up and
other early-stage ventures, respectively, Australia and Italy spent 0.87%
and 0.45%, respectively. Similarly for venture capital investment allo­
cated to later-stage ventures: the United States, for example, spent an
average of 21.4% of GDP while France and Germany, 2.3% and 1.4%,
respectively.
Second, this study offers a new angle to the extant literature on
sustainable development (e.g., Bradač & Crnogaj, 2020; He, Nazari,
Zhang, & Cai, 2020; Hörisch, Kollat, & Brieger, 2017) by studying
whether socially-oriented businesses and their counterparts exhibit
distinct patterns of financing. This discussion is complemented with an
analysis of the main drivers of social entrepreneurship. To this end, in­
formation from the GEM APS 2015, special report on social entrepre­
neurship (Bosma, Schøtt, Terjesen, & Kew, 2016) is also used.
Specifically, the survey broadly defines social entrepreneurship as any
kind of activity, organization or initiative that has a specific social,
environmental or community objective.
This study is organized as follows. Section 2 states the hypotheses of
interest in the context of the extant literature, while Section 3 describes
the data and estimation methods. Section 4 in turn presents and dis­
cusses the empirical findings on entrepreneurial finance and social
entrepreneurship. Lastly, Section 5 closes by summarizing the main
findings and discussing their policy implications.
2
International Review of Financial Analysis 75 (2021) 101722
2. Hypotheses and previous literature
Entrepreneurial finance encompasses a wide array of topics, such as
fundraising, investing, staging of investment, syndication, financial
contracting, monitoring, selling companies in initial public offerings and
acquisitions, and dealing with failed investments (Cumming et al., 2019;
Cumming & Johan, 2017). This article focuses on the fundraising stage.
To this end, three hypotheses, aimed at exploring new themes not
covered by recent research on entrepreneurial finance and social
entrepreneurship, are formulated:
H1. : Demographics, such as age, gender, educational level; entrepre­
neurial experience in helping to start a business or providing financing
as a business angel; business features, such as export intensity and
growth expectations; and, phase of economic development are the main
drivers of entrepreneurial financing. In turn, the novelty of the product/
service offered, financial barriers and investor/property rights protec­
tion also play a role, particularly in determining the amount of self-
investment.
In regards to H1, Frešer and Tominc (2018) studied whether the
degree of innovativeness associated to products, markets, and technol­
ogy affects the type and number of financial sources. Nofsinger and
Wang (2011) in turn analyzed the extent to which institutional investors
rely on entrepreneurial experience to run a business and on the quality
of investor protection to reduce moral hazard issues. In related research,
Korosteleva and Mickiewicz (2011) investigated whether the amount of
money used in new ventures is affected by the quality of the legal sys­
tem, government intervention in the financial sector, and by net inter­
national capital flows.
Our analysis of entrepreneurial finance builds upon these three ar­
ticles by considering the degree of business internationalization and
growth potential, along with the phase of economic development, de­
gree of investor and property rights protection, financial development,
among other factors that make up the entrepreneurial ecosystem. In
addition, our research offers a comprehensive perspective of start-up
financing choices by modeling the determinants of the amount needed
to start a new business, of the number of external sources of financing
used, and of the share of self-investment provided to the new business.
H2. : In countries with a high level of social trust, early-stage entre­
preneurs are more likely to obtain external financing.
Ding et al. (2015) argue that social trust influences angel investment
because this relies on information exchange and collaboration between
individuals, and it also operates as a mechanism of effective sanctioning.
Building upon this research, we conjecture that more social trust may
make it easier for early-stage entrepreneurs to obtain external financing.
In order to test this hypothesis, we define a binary variable that takes
on the value of 1 if the country in question has a high level of social trust;
and 0 otherwise. This threshold of social trust is defined in such a way
that countries at all stages of economic development are represented.
Furthermore, instead of adding this binary variable as an additional
covariate in the statistical model of external financing, we use it as a
treatment variable. In fact, as we will see, a matching technique is more
powerful than standard regression analysis in revealing the relevance of
social trust in obtaining external financing.
H3. : New and established socially-oriented businesses exhibit distinct
patterns of financing.
To our knowledge, the testing of H3 is new to the literature. Indeed,
recent research on this area has highlighted other aspects of the social
entrepreneurial process, such as human capital, cultural differences, and
environmental awareness. For instance, Estrin, Mickiewicz, and Stephan
(2016) analyzed the role of specific and general entrepreneurial human
capital in social and commercial entrepreneurship; while Bradač and
Crnogaj (2020) studied whether cultural differences between North­
western and Southeastern European countries affect social
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

entrepreneurship, in terms of innovation and market activity. The arti­ Table 1

cles by Fernandez-Laviada, Lopez-Gutierrez, and San Martin (2020) and Descriptive statistics of GEM Adult Population Survey (APS) 2015.
Fernandez-Laviada, Lopez-Gutierrez, and Perez (2020) in turn focused (a). Number of individuals by region and economic development level.
on the impact of economic development, societal values and entrepre­
Region Factor Efficiency Innovation Total
neurial environment on the likelihood of becoming a social entrepre­ driven driven driven
neur. On the other hand, Hoogendoorn, van der Zwan, and Thurik
Africa 9436 9704 – 19,140
(2020) explored the impact of environmental value creation on start-up Asia & Oceania 2106 2600 4055 8761
innovation when controlling for the regulatory context, and concluded Europe 10,647 40,368 75,617 126,632
that greener new ventures are more likely to engage in product and Latin America & – 20,187 – 20,187
process innovation. We build upon these articles by adding the financial Caribbean
North America 6561 6561
perspective of social initiatives. – –
Total 22,189 72,859 86,233 181,281

(b). Individuals involved in TEA by geographic region

3. Methodology
Involved in TEA
3.1. Data Region No Yes Total

Africa 15,434 3706 19,140

This study is based on the 2015 GEM Adult Population Survey (APS), Asia & Oceania 7384 1377 8761
which includes the special topics of finance and social entrepreneurship. Europe 114,232 12,400 126,632
The 2015 APS comprises 181,126 individuals of 62 countries (see Ap­ Latin America & the Caribbean 16,318 3869 20,187
North America 5814 747 6561
pendix 1A). Following the World Economic Forum (WEF) classification
Total 159,182 22,099 181,281
of economic development level, GEM divides countries into factor-, ef­
ficiency- and innovation-driven economies. Specifically, factor-driven (c). Individuals involved in TEA by geographic region and economic sector

economies are the least developed, and are dominated by subsistence TEA: Firm type
agriculture and extraction businesses, with a heavy reliance on unskilled Region Extractive Transforming Services Consumer Total
labor and natural resources (e.g., India, Iran, Senegal, and Vietnam). oriented
Efficiency-driven economies in turn are increasingly competitive, with Africa 688 970 206 1688 3552
more-efficient production processes and increased product quality (e.g., Asia & 353 870 423 3490 5136
South Africa, China, Indonesia, Brazil, and Chile); while innovation- Oceania
driven economies are the most developed, where businesses are more Europe 440 1258 1539 2801 6038
Latin 219 1283 721 3898 6121
knowledge-intensive and the service sector expands (e.g., Australia,
America &
Israel, South Korea, Germany, the United Kingdom and the United the
States). The number of observations by geographic region and economic Caribbean
development level are provided in Table 1(a). (See Table 1.) North 26 133 203 330 692
The focus of this study is Total early-stage Entrepreneurial Activity America
Total 1726 4514 3092 12,207 21,539
(TEA). According to GEM, TEA encompasses start-ups that have not paid
wages in the last 3 months (i.e., nascent businesses) and owner- Notes: (1) TEA stands for Total Early-Stage Entrepreneurial Activity. (2)
managers of a new firm (i.e., less than 3.5 years old). Established busi­ Extractive: agriculture, forestry & fishing, mining & quarrying (ISIC4: divisions
nesses in turn are those that are more than 3.5 years old. Descriptive 100–990); Transforming: manufacturing (processing), construction, and
wholesale trade (ISIC4: divisions 1100–4690); Services: professional and busi­
statistics are presented in Tables 1(b) and 1(c). In particular, Table 1(b)
ness services (ISIC4: divisions 5800–8299); Consumer oriented: trading activ­
shows the frequencies of individuals involved in TEA and of those
ities to the final consumer (retail, accommodation & food service) and personal
running established businesses by geographic region. As can be seen services (ISIC4: divisions 4711–4799; 5510–5630; 8510–9609).
from the figures, around 12% of the sample (22,099 individuals) cor­
responds to early-stage entrepreneurs, with Africa and Latin America &
entrepreneurs in Sectors 2 and 4 (73%). With respect to the motive to get
the Caribbean being the regions where individuals are comparatively
involved in TEA, 68% of females do due to business opportunity and
more involved in TEA (19% each).
around 30% due to necessity. Males’ ventures in turn are comparatively
Table 1(c) in turn classifies early-stage businesses by region and
more opportunity- than necessity-driven (75% versus 23%). Now,
economic sector: (1) Extractive: agriculture, forestry & fishing, mining &
regardless of gender, entrepreneurial activity occurs most frequently
quarrying. (2) Transforming: manufacturing (processing), construction,
between the ages of 25 and 44 (57% of females and 56% of males). This
and wholesale trade. (3) Services: professional and business services. (4)
could be attributed to the fact that these individuals have already
Consumer oriented: trading activities to the final consumer (retail, ac­
developed their skills and knowledge through education and work
commodation & food service) and personal services. In the full sample,
experience, and have become more confident in their own abilities. In
early-stage businesses in Sectors (1), (2), (3), and (4) represent,
addition, it is also likely that they rely on networks, personal savings and
respectively, 8%, 21%, 14%, and 57%. Due to its high dependence on
access to financing (Daniels et al., 2016).
natural resources, Africa exhibits the highest proportion of early-stage
Lastly, Table 1(d) presents descriptive statistics of sources of entre­
extractive businesses (19%), followed by Asia & Oceania and Europe
preneurial finance for early-stage and established businesses (see Ap­
(7% each). On the other hand, most of the new businesses in Europe and
pendix 2) by phase of economic development. As can be seen, at any
North America concentrate in Sectors (3) and (4)—71% and 77%,
business stage, innovation-driven economies engage in businesses of
respectively, whereas those in Africa, Asia & Oceania, and Latin America
greater scope which demand more financing, rely less on resources from
& the Caribbean in Sectors (2) and (4)—75%, 85%, and 85%, respec­
family members and friends/neighbors (i.e., love money) and more on
tively. A similar pattern is found if early-stage businesses are broken
private investors/venture capitalists and government support, relative
down by economic sector and phase of economic development (not re­
to factor- and efficiency-driven economies. Note should be taken,
ported). This is not surprising given the classification reported in Ap­
however, of the fact that the empirical distributions of total investment,
pendix 1A.
self-investment, and of external sources of financing are highly skewed,
On the other hand, when looking at sector choices by gender (not
so that sample means differ from sample medians by a great extent. For
reported), the data shows that female early-stage entrepreneurs are
instance, the medians of start-ups’ total investment for factor-,
mainly concentrated in Sector 4 (69%) while male early-stage

3
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

Table 1
Descriptive statistics (continued)
(a). Funding sources by business stage and economic development level

Factor driven Efficiency driven Innovation driven

Variable Obs. Mean Obs. Mean Obs. Mean

Total required (USD) 3946 59,457 8620 83,591 3877 615,431

Own money (% provided) 3574 62.3 7767 75.6 3316 72.7
Early-stage Family members (% using) 3927 48.1 9324 40.8 4710 22.6
Friends/Neighbors (% using) 3665 20.4 9321 10.0 4748 7.3
Employer/colleagues (% using) 3593 6.4 9317 8.4 4744 9.4
Banks (% using) 3651 22.6 9308 28.3 4704 28.0
Private investors/VC (% using) 3596 9.3 9282 8.6 4704 13.2
Government (% using) 3585 17.2 9282 13.2 4684 21.3
Crowd-funding (% using) 2034 0.9 6737 5.0 4015 5.4
Total required (USD) 891 4103 972 25,221 1153 270,315
Own money (% provided) 842 63.5 838 78.2 763 74.6
Established Family members (% using) 900 39.8 1139 38.2 1629 18.4
Friends/Neighbors (% using) 866 12.6 1138 8.7 1633 2.6
Employer/colleagues (% using) 865 7.1 1138 6.9 1636 4.1
Banks (% using) 873 21.5 1143 32.4 1637 35.6
Private investors/VC (% using) 860 5.6 1137 6.9 1631 4.8
Government (% using) 865 10.1 1136 9.7 1632 14.0
Crowd-funding (% using) 187 2.1 647 2.9 1360 1.5

Note: Early-stage entrepreneurs are made up of nascent entrepreneurs, those who have taken steps to start a new business but have not yet paid salaries or wages for
more than three months; and, new entrepreneurs, those who have been in business for between 3 and 42 months (3.5 years). Established businesses are those that are
3.5 years old or older.

efficiency- and innovation-driven economies are $940, $3010, and
$13,986, respectively, in contrast with the means reported in Table 1
(d)—$59,457, $83,591, and $615,431, respectively. This aspect is taken
into consideration in the statistical modeling of Section 4.
It should be noted that the information collected by the GEM APS
2015 does not contain details on the amount of money received or ex­
pected to be received from the alternative financing sources detailed in
Table 1(d). In particular, for established business that received money
from banks, for instance, we do not know the amount of outstanding
debt; nor figures on profits and/or net worth. This is in contrast to da­
tabases for the United States, such as the Kauffman Firm Survey (e.g.,
Robb & Watson, 2012).
3.2. Estimation methods
Given the nature of GEM APS information, which combines cate­
gorical and continuous variables, the methodology for testing the hy­
potheses of Section 2 will involve two-sample Wilcoxon rank-sum test
and a multi-sample generalization of it, the Kruskal-Wallis test (e.g.,
Newbold, Carlson, & Thorne, 2020, ch. 15), robust regression (e.g.,
Andersen, 2008), bivariate probit and ordered logistic models (e.g.,
Greene, 2018, ch. 17), and matching techniques (e.g., Cameron &
Trivedi, 2005, ch. 25).
4. Results and discussion
4.1. Sources of entrepreneurial finance
4.1.1. Determinants of start-up investment, external and self-financing
This section concentrates on the amount of money needed to start a
new business, the percentage of the entrepreneur’s own money invested,
and the number of external sources of financing obtained or expected to
be obtained. The latter is constructed as the summation of the binary
variables described in Appendix 2. For instance, if an early-stage
entrepreneur answers yes (= 1) to three of these questions (.e.g., fam­
ily members, banks and government program), the number of external
sources of financing provided or expected is recorded as 3. For illus­
trative purposes, Table 2 presents descriptive statistics of this variable
by economic development level. As can be seen, regardless of the latter,
around 87–90% of the individuals involved in TEA utilize 0–2 external
sources of financing. This means that only about 10–13% of early-stage
entrepreneurs rely on 3–7 alternative sources of external financing.
More specifically, the most frequent number of external sources of
financing is 1 among factor- and efficiency-driven economics (42% and
44%, respectively), and 0 among innovation-driven economies (45%).
Table 3 presents equality-of-populations rank tests of the percentage
of own funds provided and the number of external sources of financing
with respect to industry sector, economic development level, educa­
tional level, household-income level, expected jobs in the 5 next years,
investor protection level, new product & few competitors, latest tech­
nology, gender, and improvement-driven opportunity (IDO) entrepre­
neur (i.e., that seeking to improve his/her situation either through
increased independence or through increased income, versus main­
taining his/her income).1 Investor protection corresponds to the
2014–2015 strength of investor protection index recorded by The Global
Competitiveness Index Historical Dataset 2005–2014 on a scale of
0 (little to none) to 10 (best), https://reports.weforum.org/. This index
in turn is an average of 3 indices: the extent of disclosure index (trans­
parency of transactions), the extent of director liability index, and the
ease of shareholder suit index.
The results based on Kruskal-Wallis (multi-sample) and Wilcoxon
(two-sample) rank-sum tests show that early-stage entrepreneurs
seeking service and consumer-oriented ventures tend to provide a higher
percentage of their own money. This is also true for early-stage entre­
preneurs located in efficiency- and innovation-driven economies, early-
stage entrepreneurs with some secondary education and secondary de­
gree, early-stage entrepreneurs with relatively low expectations of job
creation (no jobs/1–5 jobs in the next five years), early-stage entrepre­
neurs using the latest technology, female early-stage entrepreneurs, and
1
Related research has focused on the impact of social media activity (e.g.,
Facebook, LinkedIn, and Twitter) on the likelihood of obtaining external
financing. For instance, Nigam, Benetti, and Johan (2020) analyzed active and
failed digital Indian startups during 2014− 2017, and concluded that
networking signals (e.g., a degree from an elite educational institution) and
digital signals (e.g., active presence on social media sites) were more relevant to
obtaining venture capital financing than traditional quality signals (e.g., prior
industry experience). GEM APS 2015 does not contain information on social
media activity, but only on whether individuals received or expect to receive
financing from crowd-funding platforms.

4
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

Table 2
N of external sources of financing for early-stage entrepreneurs by economic development level.
◦

Factor driven Efficiency driven Innovation driven

N. Freq. % Cum. Freq. % Cum. Freq. % Cum.

◦

0 1082 27.37 27.37 2864 30.46 30.46 2138 44.63 44.63

1 1668 42.2 69.57 4129 43.91 74.36 1319 27.54 72.17
2 793 20.06 89.63 1447 15.39 89.75 715 14.93 87.1
3 284 7.18 96.81 581 6.18 95.93 371 7.75 94.84
4 84 2.12 98.94 229 2.44 98.36 145 3.03 97.87
5 29 0.73 99.67 107 1.14 99.5 59 1.23 99.1
6 11 0.28 99.95 30 0.32 99.82 24 0.5 99.6
7 2 0.05 100 17 0.18 100 19 0.4 100
Total 3953 100 9404 100 4790 100

Note: economic development level is as defined in Appendix 1A.

non-IDO early-stage entrepreneurs.
Meanwhile, the relationship between the percentages of own money
provided to the new business and investor protection is non-monotonic.
Indeed, for countries with an investor protection less than 6, the rela­
tionship between the two variables is increasing, on average. That is, in
countries with low or intermediate levels of investor protection, entre­
preneurs are less likely to obtain external financing, either from informal
investors or institutions (e.g., Nofsinger & Wang, 2011). However, as
investor protection reaches high levels, the average percentage of own
resources decreases. Indeed, the data shows (not reported) that entre­
preneurs located in countries with a level of investor protection above 6
rely more on banks, private investors/venture capital and government
support.
Regarding external sources of financing, a slightly greater number of
them is received or expected by new businesses in the extractive and
transforming sectors, in factor- and efficiency-driven economies, new
businesses introducing new products and facing few competitors, new
businesses having high expectations of job creation (6–19/20+ jobs in
the next five years), new business run by entrepreneurs with post-
secondary education and graduate experience, and by new business in
countries with increasing investor protection. It should be noted, how­
ever, that the mean values of external sources of financing reported in
Table 3 range from 0.8 (no jobs expected in the next 5 years) to 1.7 (20+
jobs expected in the next 5 years), as around 87–90% of the individuals
involved in TEA utilize only 0–2 external sources of financing (Table 2).
On the other hand, household income level is not related to either the
percentage of own funds provided or the number of external sources of
financing.
In order to complement the above discussion, Appendix 1B presents
figures of the percentage of own funds provided and of the number of
external sources of financing (means)—both from GEM, strength of
investor protection (already defined), financial development (on a scale
of 0–1, IMF Financial Development Index Database 2015 (https://data.
imf.org/), property rights (on a scale of 1–7, Global Competitiveness
Index Historical Dataset 2005–2014), level of social trust (World Values
Survey, WVS, http://www.worldvaluessurvey.org/wvs.jsp),2 and usage
of money from family, friends/neighbors, private investor (PI)/venture
capital (VC) and government, by country and legal origin. The latter
classification is based on La Porta, Lopez-de-Silanes, and Schleifer
(2008): common law English origin (13 countries; 2876 observations),
civil law German origin (13 countries; 1514 observations), civil law
Scandinavian origin (3 countries; 270 observations), and civil law
French origin (31 countries; 9997 observations). These four external
2
It measures the percentage of individuals in each country who believe that
others can be trusted, on the basis of the following question (Variable A165 in
WVS):” In general, do you think that most people can be trusted, or one can’t be
too careful when dealing with people?” Examples of empirical studies using
WVS data can be found in World Values Research, http://www.worldval
uessurvey.org/WVSPublicationsPapers.jsp
sources of financing are chosen because they exhibit statistical differ­
ences across groups, according to the Kruskal-Wallis rank-sum and joint
equality F tests.
In this respect, La Porta, Lopez-de-Silanes, Schleifer, and Vishny
(1997) showed that legal rules that protect investors vary systematically
across legal origins, with common law countries being more protective
of outside investors than civil law countries, particularly French civil
law ones. Specifically, civil law is usually associated with more gov­
ernment ownership and regulation— and, hence, with greater corrup­
tion, a larger informal economy, and higher unemployment— than
common law. Furthermore, La Porta et al. (2008) argue that a strong
conception of legal origin is that common law seeks to support private
market initiatives, whereas civil law seeks to replace them with state-
desired allocations. That is, common law is dispute-resolving while
civil law is policy-implementing. Moreover, in La Porta et al.’s view,
legal origin is central to the understanding of different types of capi­
talism. For instance, legal-origin heritage may explain heavily over­
regulated economies of developing countries.
Appendix 1B shows that the country groups of English- and Scandi­
navian- legal origin display on average higher indices of investor and
property rights protection and financial development, relative to those
of French- and German-legal origin. However, the average percentage of
own funds provided and the average number of external sources of
financing used do not differ substantially across these four groups: (73%,
1.13)—English, (70%, 1.12)—German, (69%, 0.98)—Scandinavian, and
(71%, 1.10)—French. It should be noted, nevertheless, that there may be
substantial variation within groups, as, for instance in the English-legal
origin group, Thailand’s average percentage of own funds is 85%.
As to external sources of financing, on average, individuals located in
countries of Scandinavian- legal origin received or expect to receive
money less frequently from family and friends/neighbors (17% and 5%,
respectively) and more frequently from PI/VC and government (14%
and 21%, respectively), relative to the remaining groups. If Norway is
excluded, the mean percentages associated to family, friends/neighbors,
PI/VC and government increase to 24%, 7%, 16%, and 24%, respec­
tively, but the conclusion still holds. On the other hand, countries of
French-legal origin are the ones that, on average, rely the most on
friends/neighbors (39%) and the least on PI/VC and government (10%
and 14%, respectively). However, as remarked earlier, there is sub­
stantial intra-group heterogeneity. For instance, 20% of Greek early-
stage entrepreneurs received or expect to receive money from PI/VC,
whereas 46% of them received or expect to receive money from their
government.
With regards to social trust, countries of Scandinavian- legal origin
exhibit the highest average level of social trust (65%), followed by those
of English-legal origin (30%), of German-legal origin (26%), and of
French-legal origin (19%). However, once again, intra-group variation
may be substantial. For instance, among countries of English-legal
origin, Australia exhibits considerable more social trust (54%) than
average whereas Malaysia, much less (9%). Extreme values of social
trust are also observed among countries of French-legal origin, e.g.,

5
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

Table 3 Table 3 (continued )

Equality-of-populations rank tests (b) Wilcoxon rank-sum (Mann-Whitney) test
(a) Kruskal-Wallis
Percentage of self- No. of external sources of
Percentage of self- No. of external sources of investment financing
investment financing
New product & Obs. Mean Mean Obs. Mean Mean
Sector Obs. Mean Mean Obs. Mean Mean few competitors rank own rank sources
rank own rank sources
No 7257 7511 73% 9121 9034 1.10
Extractive 1162 6377 65% 1384 9321 1.20 Yes 7400 7151 70% 9026 9114 1.13
Transforming 2983 6872 69% 3637 8963 1.15 p-value = 0.000 p-value =0.275
Services 1837 7186 72% 2449 8616 1.12
Note: Investor protection corresponds to the 2014–2015 strength of investor
Consumer 8382 7409 74% 10,219 8797 1.08
oriented protection index recorded by The Global Competitiveness Index Historical
p-value = 0.000 p-value =0.000 Dataset 2005–2014 on a scale of 0 (little to none) to 10 (best), https://reports.
Development Obs. Mean Mean Obs. Mean Mean weforum.org/.
level rank own rank sources
Factor 3574 6267 65% 3953 9613 1.18
Netherlands (66%) and Philippines (3%).
Efficiency 7767 7751 76% 9404 9177 1.12
Innovation 3316 7479 73% 4790 8413 1.05 Interestingly, the data shows that social trust in innovation-driven
p-value = 0.000 p-value =0.000 economies is, on average, larger (33%) than in factor- and efficiency-
Education Obs. Mean Mean Obs. Mean Mean driven ones (25% and 20%, respectively). By contrast, if one considers
rank own rank sources
median values, factor- and efficiency-driven economies exhibit an
None 2416 6374 64% 2564 8424 0.96
Some secondary 2085 7578 74% 2610 8851 1.05
almost identical level of social trust (around 14%), whereas innovation-
Secondary 4823 7630 75% 6104 9174 1.15 driven economies double such level (30%). It is also worth indicating
degree that social trust is only weakly correlated with investor protection
Post-secondary 4423 7303 72% 5631 9146 1.16 (− 0.04), but much more correlated with financial development and the
Graduate 812 6995 69% 1086 8969 1.16
strength of property rights protection (0.50 and 0.54, respectively).
experience
p-value = 0.000 p-value =0.000 In sum, the above results indicate there may be substantial hetero­
Household Mean Mean Obs. Mean Mean geneity across countries of the same legal origin, with respect to the
income rank own rank sources variables of interest of this study.3 Therefore, the statistical models
Lowest 33- 3842 6689 71% 4669 8182 1.16 described below control for economic development level and geographic
percentile
Middle 33- 4258 6834 73% 5167 8032 1.11
region rather than for legal origin. In particular, the focus of analysis is
percentile the determinants of the amount of money needed to start a new business
Upper 33- 5401 6721 71% 6337 8064 1.14 (measured in 2015 international dollars, i.e., PPP), the number of
percentile external sources of financing, and the percentage of self-investment.
p-value = 0.125 p-value =0.271
Given that the series of amount of money needed to start and the per­
Expected jobs in Obs. Mean Mean Obs. Mean Mean
5 years rank own rank sources centage of self-investment exhibit skewness and excess kurtosis, the
No job 1611 7635 83% 2010 6219 0.78 statistical estimation is based on robust-regression models with Huber
1–5 jobs 7384 6568 73% 8562 7276 1.03 weights/bi-weights.4 The number of external sources of financing in
6–19 jobs 2512 5494 63% 2944 8322 1.36 turn takes on discrete values, so a Poisson regression model is used.
20+ jobs 1228 5293 60% 1518 9025 1.65
p-value = 0.000 p-value = 0.000
The independent variables in these regression models (Tables 4(a)-
Investor Obs. Mean Mean Obs. Mean Mean (c)) include previous entrepreneurial experience in helping to start a
protection rank own rank sources business (GEM suacts) and as a business angel (GEM busang); business
≤4.0 970 6495 61% 2923 8861 1.03 features—export intensity, expected number of jobs in the next five
>4.0 and ≤5.0 3119 6028 74% 4513 8686 1.04
years, number of owners, new product/few competitors, use of latest
>5.0 and ≤6.0 3178 6104 77% 4454 9093 1.10
>6.0 and ≤7.0 1947 6317 73% 2716 9426 1.18 technology, improvement driven opportunity (IDO); industry sec­
>7.0 3232 6405 71% 3541 9432 1.24 tor—extractive, transforming, services, and consumer oriented; institu­
p-value = 0.000 p-value =0.000 tional environment— strength of investor protection and property
(b) Wilcoxon rank-sum (Mann-Whitney) test rights, and financial development; and, as control variables, de­
mographic features—entrepreneur’s age, gender, work status, and
Percentage of self- No. of external sources of
investment financing educational level; geographic region—Africa, Asia & Oceania, Europe,
Latin America & the Caribbean, and North America; and economic
New product & Obs. Mean Mean Obs. Mean Mean
few competitors rank own rank sources development level—factor-, efficiency-, and innovation-driven.
The choice of the above variables is according to Nofsinger and Wang
No 10,615 7329 72% 13,262 8867 1.05
Yes 4042 7329 72% 4885 9630 1.28
(2011)’s study on entrepreneurial finance based on GEM 2003 APS,
p-value = 0.996 p-value =0.000 Daniels et al. (2016)’s discussion on GEM entrepreneurial finance spe­
Latest Obs. Mean Mean Obs. Mean Mean cial topic, and Amoros, Etchebarne, Zapata, and Felzensztein (2016)’s
technology rank own rank sources
No 11,964 7287 71% 14,937 8971 1.09
Yes 2693 7515 73% 3210 9554 1.21
3
p-value = 0.006 p-value =0.000 Nofsinger and Wang (2011) also refer to legal origin in their analysis.
Gender Obs. Mean Mean Obs. Mean Mean However, their sample of 27 countries is much more homogeneous than ours.
rank own rank sources 4
At the center of the distribution, the Huber weight function behaves like the
Female 6184 7678 75% 9652 7113 1.04 mean and the least squares objective function associated with it, so that ob­
Male 8473 7074 70% 8495 11,302 1.16
servations are given equal weight. At the extremes, by contrast the Huber
p-value = 0.000 p-value =0.000
weight function behaves like the median and the least absolute values objective
Improvement- Obs. Mean Mean Obs. Mean Mean
driven rank own rank sources function associated with it, so that observations are down-weighted as they get
opportunity farther out on the tails. The major difference between the Huber weight and bi-
weight functions occurs at the extremes, where the latter is somehow more
resistant to outliers (Andersen, 2008, pages 18–19).

6
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

Table 4
Determinants of entrepreneurial financing.
(a) Dependent variable: Amount of money needed to start a business (TEA) (International dollars)

Covariates Coefficient. Std. Err. t P-value

Age 15.1 6.9 2.19 0.03

Male 2295.7 166.5 13.79 0.00
Education
Some secondary 314.4 302.3 1.04 0.30
Secondary degree 1535.9 263.7 5.82 0.00
Post-secondary 2512.7 273.5 9.19 0.00
Grad experience 2553.9 413.7 6.17 0.00
Export intensity
More than 75% 523.3 462.3 1.13 0.26
25 to 75% 2862.7 290.3 9.86 0.00
Under 25% 1868.6 192.0 9.73 0.00
Expected no. jobs in 5y
1–5 jobs 1269.4 245.1 5.18 0.00
6–19 jobs 4479.9 291.6 15.36 0.00
20+ jobs 5264.1 347.4 15.15 0.00
Number of owners 39.7 6.3 6.26 0.00
New product/few comp. 16.8 178.6 0.09 0.93
Latest technology − 521.6 213.3 − 2.45 0.01
Improvement driven 260.3 161.6 1.61 0.11
Region
Africa 27.8 596.0 0.05 0.96
Asia & Oceania 1217.8 538.1 2.26 0.02
Europe 2803.6 477.8 5.87 0.00
Latin America & the Caribbean − 1307.8 555.7 − 2.35 0.02
Industry sector
Transforming 921.3 329.5 2.80 0.01
Services 280.3 368.7 0.76 0.45
Consumer oriented 372.3 307.3 1.21 0.23
Development level
Factor-driven − 1532.8 377.8 − 4.06 0.00
Efficiency-driven 740.2 297.4 2.49 0.01
Constant 435.9 713.3 0.61 0.54
N = 12,984 Adjusted R2 = 15.1% Prob > F = 0.00

(b) Dependent variable: Number of external sources of financing (TEA)

Covariates Marginal effect Std. Err. t P-value

Age − 0.010 0.001 − 11.04 0.00

Male 0.052 0.022 2.40 0.02
Business experience 0.127 0.030 4.25 0.00
Business angel 0.367 0.055 6.68 0.00
Education
Some secondary 0.080 0.043 1.87 0.06
Secondary degree 0.065 0.037 1.75 0.08
Post-secondary 0.027 0.038 0.70 0.48
Grad experience − 0.016 0.052 − 0.31 0.76
Export intensity
More than 75% 0.306 0.068 4.46 0.00
25 to 75% 0.379 0.042 9.07 0.00
Under 25% 0.212 0.026 8.15 0.00
Expected no. jobs in 5y
1–5 jobs 0.193 0.031 6.15 0.00
6–19 jobs 0.458 0.038 12.14 0.00
20+ jobs 0.615 0.046 13.41 0.00
Number of owners 0.000 0.000 0.78 0.43
New product/few com 0.147 0.022 6.73 0.00
Latest technology 0.041 0.027 1.54 0.12
Improvement driven − 0.060 0.021 − 2.85 0.00
Region
Africa − 0.150 0.067 − 2.25 0.03
Asia & Oceania 0.053 0.065 0.81 0.42
Europe 0.008 0.059 0.14 0.89
Latin America & the Caribbean − 0.088 0.065 − 1.37 0.17
Industry sector
Transforming − 0.083 0.043 − 1.92 0.06
Services − 0.071 0.048 − 1.50 0.14
Consumer oriented − 0.117 0.041 − 2.86 0.00
Development level
Factor-driven 0.615 0.077 7.95 0.00
Efficiency-driven 0.267 0.046 5.83 0.00
Investor protection 0.059 0.008 7.24 0.00
Financial development 0.270 0.088 3.08 0.00
N = 10,817 Pseudo R2 = 3.5% Prob > χ2 (28) = 0.00

7
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

(c) Dependent variable: Percentage of self-investment (TEA)

Covariates Coefficient Std. Err. t P-value

Age 0.10 0.03 3.39 0.00

Male − 2.80 0.75 − 3.71 0.00
Business experience − 12.50 1.56 − 7.99 0.00
Education
Some secondary 8.91 1.42 6.28 0.00
Secondary degree 5.12 1.25 4.10 0.00
Post-secondary 5.07 1.29 3.95 0.00
Grad experience 3.84 1.83 2.09 0.04
Work status
Not working − 2.51 1.20 − 2.10 0.04
Retired students − 4.29 2.19 − 1.96 0.05
No formal financing 5.96 0.84 7.08 0.00
N external sources of financing
◦
− 4.33 0.34 − 12.76 0.00
Export intensity
More than 75% − 4.68 2.04 − 2.29 0.02
25 to 75% − 2.82 1.23 − 2.30 0.02
Under 25% − 3.03 0.84 − 3.61 0.00
Expected No. jobs in 5y
1–5 jobs − 7.63 1.22 − 6.26 0.00
6–19 jobs − 15.66 1.39 − 11.26 0.00
20+ jobs − 16.72 1.60 − 10.43 0.00
Region
Africa − 24.27 2.49 − 9.73 0.00
Asia & Oceania − 0.58 2.26 − 0.26 0.80
Europe − 3.52 1.99 − 1.77 0.08
Latin America & the Caribbean − 11.56 2.35 − 4.92 0.00
Industry sector
Transforming − 1.04 1.45 − 0.71 0.48
Services − 1.29 1.63 − 0.79 0.43
Consumer oriented − 0.43 1.35 − 0.32 0.75
Development level
Factor-driven 5.26 1.75 3.01 0.00
Efficiency-driven 8.38 1.42 5.91 0.00
Property rights 1.53 0.60 2.57 0.01
Constant 85.69 4.62 18.54 0.00
N = 8659 Pseudo R2 = 15.9% Prob > F = 0.00

Note: (1) Panels (a) and (c) are based on robust regression models, and (c) on a Poisson regression model. (2) In Panels (a) through (c), the baseline levels of the
categorical independent and control variables correspond to Export intensity: None; Expected N◦ jobs in 5y: No jobs; Region: North America; Industry sector:
Extractive; Development level: Innovation driven; Education: None; and Work status: Work (Full-time/Part-time).

analysis on internationalization of Chilean new businesses recorded by
GEM 2007–2013 APS. In particular, export intensity measures the de­
gree of orientation towards international markets,5 whereas new prod­
uct/few competitors and the use of the latest technology capture
business competitiveness, which in turn represent business-level infor­
mational characteristics. The expected number of jobs and the number
of owners in turn measure the relative size of the new business and its
growth prospects. On the other hand, the IDO binary variable captures
the fact that IDO entrepreneurs generally demand higher financing re­
quirements (Daniels et al., 2016). Greater strength of investor protection
and property rights, and more financial development in turn are deemed
to mitigate informational asymmetries involved in external financing.
In order to model the percentage of self-investment (Table 4c), one
additional independent variable is considered: No formal financing,
which equals 1 if an entrepreneur has not received or expects to receive
financing from banks, private investors/VC, government or crowd-
funding platforms. Forty-one percent of the sampled early-stage entre­
preneurs meet this condition, of which 35% relies or will rely on one
informal source of financing and around 6.4%, on 2–3 of such sources.6
The aim of constructing this dichotomous variable is to capture financial
5
Export intensity has been found to be a key determinant of business
collaboration (e.g., Hastings and Anwar 2019).
6
This definition encompasses a more restrictive one which imposes that
early-entrepreneurs necessarily rely on, at least, one informal source of external
financing (i.e., family, friends/neighbors and/or employers/colleagues).
Imposing such a constraint leaves us with only 17% of the sampled early-stage
entrepreneurs.
barriers to start-ups—e.g., those turned down by financial institutions
and/or private investors (Roper & Scott, 2009), as entrepreneurs who
rely only on relatives and/or acquaintances’ financial support are ex­
pected to provide a greater percentage of their own money to a new
business.
In Panels (a) through (c) of Table 4, the baseline levels of the cate­
gorical independent and control variables correspond to Export in­
tensity: None, Expected No. jobs in 5y: No jobs; Region: North America;
Industry sector: Extractive; Development level: Innovation driven; Ed­
ucation: None; and Work status: Work (Full-time/Part-time). Contin­
uous variables are age (years), number of owners, investor protection,
and property rights, whereas binary variables are male, new product/
few competitors, latest technology, improvement driven, and no formal
financing.
As Panel (a) shows, except for new product/few competitors and
improvement-driven, all of the independent and control variables are
statistically relevant to explain the amount of money to start a new
business. For instance, male entrepreneurs are expected to require
$2296 more (measured in international dollars) than female counter­
parts, whereas entrepreneurs who expect 20+ jobs in the next five years
require $5264 more than those who expect none. Moreover, with the
exception of Latin American & the Caribbean entrepreneurs, those
located in Asia & Oceania and Europe need a greater amount of money
to start their businesses, relative to North American entrepreneurs.
The estimation results also show that funding requirements are
increasing with age ($15 additional per year), level of education (with
the exception of some secondary), number of business owners ($40
additional per business owner), and degree of international orientation
of 25–75% or under 25% (relative to a home-based business)—$2863

8
V. Fernandez
and $1869, respectively. By contrast, the use of the latest technology
may decrease funding needs by around $522. On the other hand, a new
business in the transforming sector may require $920 more relative to
one in the extractive sector, whereas businesses in factor- and efficiency-
driven economies demand $1533 less and $740 more, respectively,
relative to those located in innovation-driven economies.
Panel (b) in turn shows that the number of external sources of
financing is increasing in the number of expected jobs in 5 years, and
positively associated with gender (although very marginally), having
business experience, having been a business angel, export intensity, the
introduction of a new product with few competitors, and with investor
protection and financial development. For instance, having provided
funds and helped to start a new business in the past increase the number
of external sources of financing by 0.5 (= 0.13 + 0.37), suggesting that
entrepreneurs with past experience as informal investors and with
managerial skills are more compelling about the suitability of their new
business. This is in agreement with the informational signaling hy­
pothesis referred to in the Introduction.
The estimation results also indicate that transforming and consumer-
oriented businesses demand slightly fewer external sources of financing
than extractive ones (− 0.08 and − 0.12, respectively); and, that busi­
nesses located in factor- and efficiency-driven economies require more
external sources than those located in innovation-driven ones (around
0.62 and 0.27, respectively). This is in accordance with the evidence
shown in Table 2, in that the frequency of zero external sources of
financing is highest in innovation-driven economies. The statistical
model also suggests that older entrepreneurs and those seeking
improvement-driven opportunities may find it slightly harder to obtain
external financing, possibly due to higher informational asymmetries.
Now, the reason why the above marginal effects are small is because
around 73% of the sampled sources of external financing are concen­
trated on the 0–1 range. Hence, in order to offer more intuitive results,
we discretize the dependent variable as follows: Level 1 = 0 external
sources; Level 2 = 1–2 external sources, and Level 3 = More than 2
external sources, and concentrate on the probability of belonging to
each level. In the sample, the frequencies associated to these levels are
34%, 56%, and 11%, respectively. We conduct this exercise by fitting an
ordered logistic model to this 3-level variable, where the explanatory
variables are as earlier. Selected marginal effects are reported in Ap­
pendix 1C.
For instance, having been a business angel reduces the likelihood of
no external sources of financing by 5.1 percentage points. Similarly, for
someone with business experience, this likelihood decreases by 12.2
percentage points. Non-negligible effects are also found for home-based
(i.e., no exports) and growth (20+ jobs in 5Y) businesses. For the former,
the likelihood of no external financing increases by 8.5 percentage
points and for the latter decreases by 11.1 percentage points. On the
other hand, business experience, growth business, and a factor-oriented
economy have large marginal impacts on the likelihood of using more
than 2 external sources of financing: 6.4, 5.9, and 11.1 percentage
points, respectively.
Lastly, Panel (c) of Table 4 shows that statistically significant de­
terminants of the percentage of self-investment are age, gender, business
experience, work status, education, financial barriers to formal
financing, number of external sources of financing, business’ expected
number of jobs and export intensity, geographic region, economic
development level, and country strength of property rights.7 Indeed, a
male entrepreneur provides around 3% points less of his own money
relative to a female counterpart; whereas an entrepreneur without ac­
cess to formal funding sources, such as banks, private/venture capital
7
An alternative estimation route would be to define external financing as
(100− self-financing) and handle the zeros by means of a Tobit model (i.e.,
Nofsinger and Wang 2011). This route provides similar numerical results (of an
opposite sign) to those reported in Table 4(c).
9
International Review of Financial Analysis 75 (2021) 101722
and government, provides around 6 percentage points more of his/her
own money.
In turn, for each additional source of external financing, self-
investment decreases by 4 percentage points. In this regard, sample
statistics show that, for instance, an entrepreneur who received or ex­
pects to receive money from one external source provides, on average,
69% of his/own money, whereas one who relies (or expects to rely) on 5
external sources provides only 55%. On the other hand, older entre­
preneurs tend to provide proportionally more personal resources: for
each additional year of age, the percentage provided increases by 0.1
percentage points.
Internationally-oriented businesses in turn demand less personal
resources than home-based ones, most likely because they face fewer
financial barriers. (This is confirmed by a Kruskal-Wallis rank test be­
tween export intensity and no formal financing, and it is congruent with
the evidence of Table 4(b)). For instance, an entrepreneur with over 75%
of his/her customers located abroad provides around 5 percentage
points less of his/her own money than one running a home-based
business. In turn business growth prospects play even a stronger role
in diminishing self-investment. Indeed, an entrepreneur who expects
20+ jobs in the next five years may provide around 17 percentage points
less than one who expects no jobs.
The estimation results also highlight that, after controlling for
several factors, entrepreneurs located in factor- and efficiency-driven
economies provide a higher share of their own money, relative to
those in innovation-driven economies: around 5 and 8 additional per­
centage points, respectively. It should be noted, however, that these
figures are explained primarily by factor- and efficiency-driven econo­
mies located in Asia (e.g., India and China—Appendix 1A), which
exhibit higher percentages of self-investment (76% and 85%, respec­
tively) than the overall means displayed in Table 1 (62% and 76%,
respectively). Reconciling this result with that obtain in Table 4(b), one
would conclude that entrepreneurs located in factor- and efficiency-
driven economies may rely on more external sources, but these may
prove proportionally modest, and, therefore, force entrepreneurs to
provide percentagewise more of their own money.
Lastly, the estimation results show that for each point-increment of
property rights protection (0–7 scale), self-investment increases by
around 2 percentage points. In other words, early-stage entrepreneurs
are willing to invest more personal resources as the strength of legal
ownership increases. In addition, it should be highlighted that, given the
nature of a robust regression, the estimated autonomous level of self-
investment (i.e., constant term) is around 86%, which is closer to a
median estimate. Indeed, if, for instance, one looks at self-investment by
economic development level, one finds that the sample medians are
60%, 100%, and 100% for factor-, efficiency-, and innovation-driven
economies, respectively, which contrast with the mean values pro­
vided in Table 1(d).
It is worth noting that the estimation approach involved in Panels
(a), (b) and (c) of Table 4 implicitly assumes a sequential decision-
making process. Indeed, (i) the first stage involves determining the
type and scope of the new business, i.e., amount of money needed
—Panel 4(a). (ii) After stage (i) is completed, the early-stage entrepre­
neur seeks external financing. (iii) Next, depending on the type and
number of external sources obtained (if any), the early-stage entrepre­
neur determines his/her investment percentage. By contrast, Nofsinger
and Wang (2011), who analyzed GEM APS 2003, followed a different
approach in which they studied the determinants of informal (e.g.,
family and friends/neighbors) and institutional (e.g., banks, private
investors/VC) financing—modeled as binary variables— by including
self-investment among the covariates. We think, however, that our
approach is more clearly structured.
Nevertheless, the sequential nature of stages (i)-(iii) is an over­
simplification because feed-back effects may take place along the whole
decision-making process. For instance, if an early-stage entrepreneur
does not receive or expect to receive any external resources, he/she may
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

decide to modify the size and nature of his/her new business; or simply Table 5
to give up on his/her entrepreneurial activity. On the other hand, in­ Seemingly unrelated bivariate probit of start-up decision and lack of formal
vestments in entrepreneurial firms may take place in several rounds, financing.
conditional on the progress of certain pre-specified targets (Cumming (a) Dependent variable: Business
et al., 2019). start
This level of complexity cannot be captured by GEM data, however. Covariate Coefficient. Std. z P-
In any case, GEM APS 2015 sheds some lights on the frequency with Err. value
which businesses are temporarily or permanently discontinued Age − 0.006 0.001 − 5.01 0.00
(EXREAS_O) for financial motives (Problems getting finance). The data Male 0.115 0.029 4.03 0.00
shows that this reason represents only 2% of the responses, with per­ Knows entrepreneur 0.119 0.029 4.07 0.00
sonal reasons and business profitability being the most frequent ones Has skills/knowledge/experience 0.192 0.038 5.05 0.00
Works (PT/FT) − 0.707 0.056 − 12.58 0.00
(33% and 19%, respectively). Post-secondary/Grad experience − 0.143 0.029 − 4.87 0.00
Industry sector
4.1.2. Bivariate decisions of business entry and informal financing Transforming − 0.037 0.059 − 0.63 0.53
In order to have a better understanding on the driving forces of Services − 0.148 0.064 − 2.32 0.02
Consumer oriented − 0.135 0.055 − 2.46 0.01
financial barriers, we estimate a bivariate probit model on the decision
Development level
of start a new business and then lack formal financing. The estimation Factor-driven 0.536 0.061 8.84 0.00
results are provided in Table 5, Panels (a)–(c). In Panel (a), the depen­ Efficiency–driven 0.073 0.046 1.59 0.11
dent variable is business start, which corresponds to the GEM bstart (b) Dependent variable: No formal financing
question: “Are you, alone or with others, currently trying to start a new
Covariate Coefficient. Std. Err. z P-value
business?” The independent variables are Knows an entrepreneur (GEM
knowent: Do you know someone personally who started a business in the Age 0.002 0.001 1.94 0.05
Male − 0.136 0.025 − 5.41 0.00
past 2 years?), Have skills/knowledge/experience (GEM suskill = 1 if
Works (PT/FT) 0.086 0.039 2.23 0.03
entrepreneur believes to have the knowledge, skill and experience Post-secondary/Grad experience 0.044 0.026 1.65 0.10
required to start a new business), Works (=1 if works part-time (PT)/full Upper income 0.098 0.025 3.93 0.00
time (FT)), and industry sector. Control variables are age, gender (=1 if Home-based business 0.142 0.026 5.41 0.00
male), educational level (post-secondary/Graduate experience), eco­ Growth business − 0.241 0.041 − 5.84 0.00
Investor protection 0.091 0.009 10.03 0.00
nomic development level, and geographic region. (See Table 5.)
Industry sector
In Panel (b), the dependent variable is no formal financing, which Transforming 0.297 0.052 5.73 0.00
has been already defined. The independent variables are Work, home- Services 0.365 0.056 6.48 0.00
based business (=1 if no customers are located abroad), growth busi­ Consumer oriented 0.270 0.048 5.61 0.00
Development level
ness (=1 if the entrepreneur expects 20+ jobs in the next 5 years),
Factor-driven − 0.394 0.058 − 6.75 0.00
country investor protection, industry sector, economic development Efficiency-driven − 0.080 0.044 − 1.84 0.07
level, and geographic region. The control variables are the same as those Includes regional dummies
in Panel (a), in addition to upper-income level (=1 if entrepreneurs
N = 11,518; ρ = − 0.13, Wald test of ρ = 0: χ2(1) = 52.9, p-value = 0.004
belongs to the 33-percentile upper income level).
(c) Marginal effects on the conditional Pr(No formal financing = 1|business start = 1)
The estimation results show that the probability of starting a new
business is positively associated with being male, knowing an entre­ Covariate dy/dx Std. Err. z P-value
preneur, and believing to have the necessary knowledge, skills and Age 0.001 0.000 1.65 0.10
experience; and, negatively associated with working part-time/full- Male − 0.050 0.010 − 5.18 0.00
time, upper income level, and having post-secondary education/grad­ Knows an entrepreneur 0.002 0.001 3.60 0.00
Has skills/knowledge/experience 0.004 0.001 4.17 0.00
uate experience. Age in turn has a very mild negative impact on the
Works (PT/FT) 0.019 0.015 1.27 0.21
start-up decision. On the other hand, entrepreneurial activity seems Post-secondary/Grad experience 0.014 0.010 1.37 0.17
stronger in factor-driven economies (relative to innovation-driven Upper income 0.038 0.010 3.93 0.00
ones), and less so in the services and consumer-oriented industry sec­ Home-based business 0.055 0.010 5.43 0.00
tors (relative to the extractive one). Growth business − 0.093 0.016 − 5.84 0.00
Investor protection 0.035 0.004 10.04 0.00
In turn, the probability of relying only on informal financing is Industry sector
positively associated with age (although very marginally), working, Transforming 0.109 0.019 5.89 0.00
having post-secondary education/graduate experience, belonging to Services 0.134 0.020 6.53 0.00
upper-income levels, having a home-based business, and with investor Consumer oriented 0.097 0.017 5.73 0.00
Development level
protection; and, it is negatively associated with being male and having a
Factor-driven − 0.137 0.021 − 6.62 0.00
growth business. In turn, industry sector and economic development Efficiency-driven − 0.029 0.017 − 1.75 0.08
level also appear to be relevant to explaining financial constraints. Region
It is worth noticing that the estimated conditional (tetrachoric) Africa 0.069 0.034 2.04 0.04
correlation between the decisions of starting a new business and Asia & Oceania 0.083 0.030 2.74 0.01
Europe − 0.006 0.027 − 0.22 0.83
depending on informal financing (ρ) is negative and equal to − 0.13. Latin America & the Caribbean 0.002 0.031 0.05 0.96
Although this figure is relatively small, it is statistically significant at any
conventional significance level. Intuitively, this result suggests that, Notes: (1) Has skills/knowledge corresponds to GEM suskills. Growth business =
1 if expects 20+ jobs in next 5 years. Upper income = 1 if belongs to GEA income
after controlling for several factors, entrepreneurship is discouraged by
upper 33% tile. (2) Marginal effects are evaluated at sample means.
the lack of institutional financing.
In order to quantify marginal effects of the covariates (dy/dx), Panel
(c) of Table 5 reports those associated with the conditional probability of income level increases it by 1.9 and 3.8 percentage points, respectively.
no formal financing given that a new business has been started, that is, On the other hand, running a home-based business increases the con­
Pr(No formal financing = 1|business start = 1). The estimation results ditional probability of no formal financing by 5.5 percentage points,
show that, for instance, being male reduces such conditional probability while running growth businesses decreases it by 9.3 percentage points.
by 5.0 percentage points, whereas working and belonging to the upper- Meanwhile, businesses belonging to the transforming, services, and

10
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

consumer-oriented sectors are more likely to rely on informal financial that remain unsatisfied by current economic or social institutions. Un­
sources (relative to the extractive one), as well as businesses located in like regular entrepreneurs, social entrepreneurs’ top priority is to create
countries with higher investor protection levels, and in Africa and Asia/ social value, while economic value creation is seen as a necessary con­
Oceania (relative to North America). The opposite occurs in factor- and dition for financial viability. Entrepreneurship in turn means that
efficiency-driven economies (relative to innovation-driven ones). exposure to market dynamics is expected to be considered a social
Altogether, the above evidence suggests that females with financial entrepreneur. This aspect distinguishes social businesses from non-profit
resources (i.e., working/earning high incomes), who run home-based organizations, whose sole goal is the social mission and any surpluses
businesses with none or moderate growth expectations, and live in generated by them are re-invested in the organization (e.g., Lepoutre,
innovation-driven economies are the most likely to get around financial Justo, Terjesen, & Bosma, 2013; Parker, 2018, pages 66–67; Bradač &
barriers. Or put another way, businesses with such characteristics are Crnogaj, 2020).
less likely to attract financing from institutional sources (e.g., banks and The GEM 2015 APS contains 34 questions on social entrepreneurship
private/venture capital) but are still viable due to the entrepreneur’s formulated via interviews to 167,793 adults in 58 economies. The GEM
financial profile. This evidence adds new information to the literature on 2016–2015 special report on this subject elaborated by Bosma et al.
female entrepreneurial propensity (e.g., Kamal & Daoud, 2018; Koel­ (2016) presents a broad and narrow measure of social entrepreneurship
linger, Minniti, & Schade, 2013; Langowitz & Minniti, 2007; Shahriar, activity. The broad measure considers individuals who are leading or
2018) and subsequent financing (e.g., Leitch, Welter, & Henry, C., starting any kind of activity, organization or initiative that has a
2018). particularly social, environmental or community objective. This might
include providing services or training to socially deprived or disabled
4.1.3. Social trust persons, activities aimed at reducing pollution or food-waste, organizing
As discussed in the Introduction, social trust is an informal mecha­ self-help groups for community action, etcetera. The narrow measure
nism that allows information transmission, cooperation, and the imposes that this activity, organization or initiative prioritizes social and
enforcement of sanctions within society. In this section, we explore the environmental value over financial value; and it operates in the market
influence of social trust on the decision of using external financing. To by producing goods and services. The narrow definition is available for
this end, we use the same series from the World Values Survey discussed 31 economies.
in Section 4.1 and reported in Appendix 1B. In particular, we would like Our analysis concentrates on the broad definition of social entre­
to find out by how much the likelihood of relying on external sources of preneurship. To this end, and in order to get a sense of the drivers of
financing increases when an early-stage entrepreneur is located in a social entrepreneurship, Table 6 shows the estimation results of fitting
country with a high level of social trust. an ordered logistic model where the dependent variable is GEM sestart:
In order to answer the above question, we created two new variables: “Are you, alone or with others, currently trying to start or currently
a binary variable that takes on the value of 1 if the early-stage entre­
preneur received or expects to receive external financing (66% of the
Table 6
sample) and 0 otherwise; and, a treatment variable that takes on the
Ordered logistic regression of activity with social, environmental or community
value of 1 if the entrepreneur’s country level of social trust is in the objective.
upper 40-percentile (≥ 28%) i.e., high level of social trust; and,
Dependent variable: sestart
0 otherwise. According to economic development level, the number of
countries which are classified as of high level of social trust is 3 (847 Covariate Odds ratio Std. Err. z P-value
observations), 4 (1966 observations), and 14 (3151 observations) Age 1.01 0.00 3.12 0.00
among the factor-, efficiency- and innovation-driven economies, Male 1.16 0.07 2.48 0.01
respectively. (The number of observation corresponds to entrepreneurs Knows an entrepreneur 1.51 0.10 6.41 0.00
Opportunity 1.18 0.07 2.71 0.01
involved in TEA). The purpose of defining a high level of social trust in Solve social problems 1.11 0.06 1.94 0.05
this way is that all levels of economic development have some repre­ Works (PT/FT) 1.21 0.11 2.19 0.03
sentation (see Appendices 1A and 1B). Post-secondary/Grad experience 0.90 0.05 − 1.69 0.09
After controlling for the same covariates as those of Table 4(b), the Investor protection 0.97 0.02 − 1.16 0.24
No formal financing 0.70 0.04 − 5.99 0.00
methodology of nearest-neighbor matching (N = 8871)—e.g., Cerulli
Home-based business 0.60 0.04 − 8.69 0.00
and Ventura (2019)— shows that the average treatment effect (ATE) of Growth business 1.54 0.12 5.32 0.00
being located in a country with a high level of social trust is an increase Extractive sector 1.58 0.14 5.08 0.00
in the likelihood of using external financing of 4.8 percentage points (p- New product/few competitors 1.60 0.09 7.96 0.00
value = 0.00). It is worth contrasting this figure with one obtained from Development level
Factor-driven 1.61 0.21 3.64 0.00
a simpler estimation route consisting of adding a high-trust binary Efficiency-driven 0.69 0.08 − 3.35 0.00
variable as a covariate in a logistic regression where the dependent Region
variable is the use of external financing and the remaining covariates are Africa 0.90 0.09 − 1.01 0.31
those of Table 4(b). This route yields a marginal effect of high trust on Asia & Oceania 0.59 0.06 − 5.28 0.00
Europe 0.53 0.06 − 5.36 0.00
the likelihood of using external financing of 2.33 percentage points,
which is statistically insignificant at the 10% level (p-value = 0.12). Notes: (1) sestart: “Are you, alone or with others, currently trying to start or
Hence, the use of matching techniques enables us to provide more currently leading any kind of activity that has a social, environmental or com­
insight into entrepreneurial financial decisions by using social trust as a munity objective?” (Appendix 2). The original variable is recoded so that 1:
treatment variable, rather than as an additional regressor. “No”, 2: “Yes, currently trying to start, 3 ‘Yes, trying to start and leading, and
4:’Yes, currently leading.” (2) Knows an entrepreneur (knowent): “Do you know
someone personally who started a business in the past 2 years?” Opportunity
4.2. Social aspects of entrepreneurial finance
(opport): “In the next six months, will there be good opportunities for starting a
business in the area where you live?” Solve social problems (nbsocent): “In my
4.2.1. Definition and determinants of social entrepreneurship country, you will often see businesses that primarily aim to solve social prob­
Social entrepreneurship is generally understood as the process of lems”. (3) The sample includes early-stage and established businesses. (4) The z-
identifying, evaluating, and exploring opportunities to create social ratio and its associated p-value correspond to those of each parameter estimate,
value by means of innovative social ventures and the use of market- β, where the odds-ratio = exp.(̂
̂ β).
based activities. The notion social refers to the fact that product and N = 9308, log pseudo-likelihood = − 5855.6, pseudo R2 = 5.0%. Wald χ2 (18) =
services developed by social entrepreneurs aim to fulfill human needs 595.5, p-value = 0.00.

11
V. Fernandez
leading any kind of activity that has a social, environmental or com­
munity objective?” (Appendix 2), which corresponds to the broad defi­
nition referred to above. In order to give it a proper order, we recode this
variable as level 1: “No”; level 2: “Yes, currently trying to start; level 3:
“Yes, trying to start and leading; and, level 4: “Yes, currently leading.” In
this regard, the data shows that 6.8% of the sampled early-stage and
established businesses are involved in social ventures (=12,067/
176,460) according to the broad definition.
The covariates of the ordered logistic model are age, gender, knows
an entrepreneur (knowent), opportunity, solve social problems, works
(part-time/full-time), post-secondary education/graduate experience,
investor protection, no formal financing, home-based business, growth
business, extractive sector (=1 if the business belongs to Agriculture,
forestry & fishing or Mining & quarrying, e.g., Table 1c), new product/
few competitors, economic development level, and geographic region.
The opportunity covariate corresponds to GEM opport: “In the next six
months, will there be good opportunities for starting a business in the
area where you live?” and, Solve social problems, to GEM nbsocent: “In my
country, you will often see businesses that primarily aim to solve social
problems”, i.e., visibility of social entrepreneurship. The choice of these
covariates is based on our previous analysis and extant literature on the
determinants of social entrepreneurship (e.g., Fernandez-Laviada,
Lopez-Gutierrez, & Perez, 2020; Fernandez-Laviada, Lopez-Gutierrez, &
San Martin, 2020; Hoogendoorn et al., 2020; Terjesen, Bosma, & Stam,
2016).
The estimation results reported in Table 6 show that the odds of
success (i.e., trying to start/leading an activity with social, environ­
mental or community objective) are greater than 1 for older entrepre­
neurs (although very marginally), males, individuals who know an
entrepreneur, individuals who see business opportunities, individuals
who think that in their country businesses are concerned with social
problems, and individuals who are currently working. In turn, the odds
ratios are less than 1 for financially-constrained individuals (i.e., those
who rely only on informal sources of external financing), for highly-
educated individuals (although this effect is not statistically significant
at the 5% level) and for individuals located in Asia & Oceania and
Europe, and in efficiency-driven economies. Investor protection (i.e.,
institutional setting) does not appear to be statistically relevant in this
case. As to the business itself, the odds exceed 1 for growth businesses,
innovative businesses which offer new products and face few competi­
tors, businesses within the extractive sector, and businesses located in
factor-driven economies. On other hand, the odds ratio is less than 1 for
home-based businesses.
Some of these findings are in agreement with previous literature. For
instance, Terjesen et al. (2016) point out that those individuals who are
employed part- or full-time or are students are more likely to start social
ventures. The same applies to males, although the gender gap in social
entrepreneurship may not be as high as in commercial entrepreneurship.
Indeed, if, for instance, one considers individuals involved in TEA, the
GEM 2015 APS shows that women located in Europe run around 40% of
socially-oriented ventures, whereas only 38% of commercial ones.
Altogether, policy implications of the results of Table 6 are that in­
dividuals in factor-oriented economies may be more involved in social
ventures that those located in more advanced economies. This has to do
with the fact that a large number of extractive ventures, which are more
socially-oriented than their counterparts,8 are located in Africa. In
addition, our results indicate that individuals who rely only on savings
and informal sources of financing, such as family, friends/neighbors,
and employer/colleagues, are less likely to get involved in activities with
a social, environmental or community objective. Hence, from a societal
viewpoint, more dedicated finance programs aimed at fostering social
8
In the sample, the percentages of extractive ventures trying to start or
currently leading a social initiative are 4.0 and 8.7, versus 2.2 and 5.3 of non-
extractive ventures.
12
International Review of Financial Analysis 75 (2021) 101722
endeavors should be in place, e.g., venture philanthropy, crowd-funding
and social impact bonds (Terjesen et al., 2016)
4.2.2. Entrepreneurial financing choices depending on social orientation
Table 7 presents two-way tables of frequencies, along with Pearson
chi-squared test of independence, of socially oriented business and
funding source —family members, friends/neighbors, employer/col­
leagues, banks, private/venture capital, government, and crowd-
funding (e.g., Table 1d). A business is defined as socially oriented if
the variable sestart equals 1, 2 or 3 (see Appendix 2). Panel (a) of Table 7
focuses on early-stage businesses whereas Panel (b), on established ones.
The testing results of Panel (a) show that at their infancy socially-
oriented businesses rely more heavily on employers/colleagues (13%
versus 7%), private investors/venture capital (17% versus 9%), gov­
ernment (25% versus 15%) and crowd-funding (10% versus 4%), and
slightly more on friends/neighbors (16% versus 11%) and banks (30%
versus 27%). There is no statistical difference regarding the use of family
funds. With regards to established businesses, Panel (b) indicates that
those socially-oriented and their counterparts display similar use of
bank- and government- funds, but the former depend more on family
(35% versus 28%), friends/neighbors (9% versus 6%), employer/col­
leagues (9% versus 5%), private investors/venture capital (9% versus
5%), and crowd-funding (6% versus 1%). Altogether, these figures
suggest that early-stage and established businesses with a social focus
appear to be more successful in obtaining financing from institutional
investors (governments, private/venture capitals) and crowd-funding
platforms.9
In order to further investigate the relationship of external financing
and social focus, we estimate average treatment effects (ATEs) for each
funding source using the socially-oriented dummy as a treatment vari­
able and several controls. The results for early-stage (TEA) businesses for
two methodologies-regression adjustment (RA) and propensity score
matching (PSM)—e.g., Cerulli & Ventura, 2019)—are reported in
Table 8. As can be seen, the greatest impact of the treatment variable is
on financial resources from private investors/venture capitalists, gov­
ernment, and crowd-funding. For instance, the likelihood for socially-
oriented businesses of using private investor/venture capitalist’s
money is 5.4 (PSM)-6.0 (RA) percentage points greater than that of their
counterparts. The largest effect is found for government financing: 6.9
(RA)-7.6 (PSM) percentage points.
5. Conclusions and policy implications
This study concentrated on financing decisions of early-stage entre­
preneurs by looking at the inter-dependence between external financing
and self-investment. In particular, a new angle to the extant literature
was offered by incorporating social trust and social awareness into
financial decisions. In addition, this study modeled the joint decisions of
starting a new business and subsequently depending only on personal
savings and financial resources from non-institutional investors, such as
family, friends/neighbors, and employers/colleagues. Several statistical
techniques were used along the analysis, such as robust regression, or­
dered logistic regression, and propensity score matching.
More specifically, our analysis of start-ups’ financial decisions
incorporated demographic characteristics, such as age, gender, and
educational level; entrepreneurial experience through helping others to
start a business and/or through providing funds as a business angel;
business features, such as growth expectations, export propensity, and
9
Now, if one looks at median values of money required to start a new
business, social orientation does not have a significant impact: USD 3250 for
socially-oriented versus USD 3181 for no socially-oriented businesses. However,
for business aged 3.5 years or over, socially-oriented ones required about one
fifth of the money required by their counterparts: USD 1106 versus USD 5551,
in median terms.
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

Table 7
Business financing choices by social-orientation.
(a) Early-stage entrepreneurs

Family Friends/Neighbors Employer/Colleagues Banks

Social No Yes Total No Yes Total No Yes Total No Yes Total

No 9326 5676 15,002 13,269 1575 14,844 13,691 1091 14,782 10,878 3922 14,800
% 62.17 37.83 100 89.39 10.61 100 92.62 7.38 100 73.5 26.5 100
Yes 1875 1084 2959 2439 451 2890 2509 363 2872 2007 856 2863
% 63.37 36.63 100 84.39 15.61 100 87.36 12.64 100 70.1 29.9 100
Total 11,201 6760 17,961 15,708 2026 17,734 16,200 1454 17,654 12,885 4778 17,663
62.36 37.64 100 88.58 11.42 100 91.76 8.24 100 72.95 27.05 100
χ2 test 1.52, p-value = 0.22 59.65, p-value = 0.00 87.99, p-value = 0.00 14.04, p-value = 0.00

Private investors/VC Government Crowd-funding

Social No Yes Total No Yes Total No Yes Total

No 13,453 1273 14,726 12,565 2133 14,698 10,447 384 10,831

% 91.36 8.64 100 85.49 14.51 100 96.45 3.55 100
Yes 2368 488 2856 2141 712 2853 1766 189 1955
% 82.91 17.09 100 75.04 24.96 100 90.33 9.67 100
Total 15,821 1761 17,582 14,706 2845 17,551 12,213 573 12,786
89.98 10.02 100 83.79 16.21 100 95.52 4.48 100
χ2 test 189.17, p-value = 0.00 191.88, p-value = 0.00 145.00, p-value = 0.00

(b) Established entrepreneurs

Family Friends/Neighbors Employer/Colleagues Banks

Social No Yes Total No Yes Total No Yes Total No Yes Total

No 2093 830 2923 2722 182 2904 2762 144 2906 1997 917 2914
% 71.6 28.4 100 93.73 6.27 100 95.04 4.96 100 68.53 31.47 100
Yes 483 262 745 664 69 733 670 63 733 515 224 739
% 64.83 35.17 100 90.59 9.41 100 91.41 8.59 100 69.69 30.31 100
Total 2576 1092 3668 3386 251 3637 3432 207 3639 2512 1141 3653
70.23 29.77 100 93.1 6.9 100 94.31 5.69 100 68.77 31.23 100
χ2 test 13.02, p-value = 0.00 9.02, p-value = 0.00 14.45, p-value = 0.00 0.37, p-value = 0.54

Private investors/VC Government Crowd-funding

Social No Yes Total No Yes Total No Yes Total

No 2756 139 2895 2565 335 2900 1933 28 1961

% 95.2 4.8 100 88.45 11.55 100 98.57 1.43 100
Yes 668 65 733 642 91 733 218 15 233
% 91.13 8.87 100 87.59 12.41 100 93.56 6.44 100
Total 3424 204 3628 3207 426 3633 2151 43 2194
94.38 5.62 100 88.27 11.73 100 98.04 1.96 100
χ2 test 18.22, p-value = 0.00 0.42, p-value = 0.52 27.2, p-value = 0.00

degree of innovativeness; and, aspects linked to the entrepreneurial
ecosystem, such as geographic location, phase of economic develop­
ment, degree of investor and property rights protection. All these aspects
were subsequently interconnected with social trust and social
entrepreneurship.
In this regard, the extant literature has shown that ongoing social
relationships may be a powerful mechanism to solve agency problems
inherent to entrepreneurial activities. For instance, our estimation re­
sults based on nearest-neighbor matching showed that being located in a
country with a high level of social trust increases the likelihood of using
external financing by around 5 percentage points.
As to socially-oriented business, our analysis showed that they are
more successful in obtaining financial resources from institutional in­
vestors. Indeed, after controlling for several factors, calculations based
on regression adjustment (RA) and propensity score matching (PSM)
showed that the likelihood for socially-oriented businesses of using
private investor/venture capitalist’s money is 5.4 (PSM)-6.0 (RA) per­
centage points greater than that of their counterparts. The largest effect
of social orientation was found for government financing: 6.9 (RA)-7.6
(PSM) percentage points.
Our statistical analysis also showed that the lack of institutional
financing is not only detrimental to social entrepreneurship but also to
any entrepreneurial activity, since decisions to start a new business and
rely exclusively on informal financing are negatively associative.
Moreover, the analysis also highlighted the fact that women are more
likely to face financial barriers after starting a new business. These re­
sults have clear public policy implications, in the sense that strength­
ening existing financing models and facilitating new ones to men and
women alike, through enhanced technology and the necessary regula­
tory framework, is crucial to new ventures, particularly those with
growth potential and propensity to export.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.

13
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

Table 8
Average Treatment Effect (ATE) of social orientation on early-stage businesses’ choice of external financing.
Regression adjustment Propensity score matching
Family Coef. Std. Err z P>z Coef. Std. Err z P>z
ATE 0.39 1.22 0.32 0.75 0.34 1.53 0.22 0.83
N = 12,513 N = 12,514
Regression adjustment Propensity score matching
Friends/Neighbors Coef. Std. Err z P>z Coef. Std. Err z P>z
ATE 2.79 0.88 3.17 0.00 3.62 1.21 2.99 0.00
N = 12,343 N = 12,343
Regression adjustment Propensity score matching
Employer/Colleagues Coef. Std. Err z P>z Coef. Std. Err z P>z
ATE 3.61 0.81 4.44 0.00 3.26 1.01 3.22 0.00
N = 12,288 N = 12,288
Regression adjustment Propensity score matching
Banks Coef. Std. Err z P>z Coef. Std. Err z P>z
ATE 3.01 1.17 2.57 0.01 2.37 1.56 1.52 0.13
N = 12,343 N = 12,315
Regression adjustment Propensity score matching
Private investors/VC Coef. Std. Err z P>z Coef. Std. Err z P>z
ATE 5.97 0.86 6.91 0.00 5.44 0.95 5.72 0.00
N = 12,245 N = 12,245
Regression adjustment Propensity score matching
Government Coef. Std. Err z P>z Coef. Std. Err z P>z
ATE 6.93 1.02 6.78 0.00 7.56 1.19 6.36 0.00
N = 12,224 N = 12,224
Regression adjustment Propensity score matching
Crowd-funding Coef. Std. Err z P>z Coef. Std. Err z P>z
ATE 5.38 0.83 6.48 0.00 6.33 1.06 5.96 0.00
N = 8796 N = 8796

Notes: (1) under the regression-adjustment methodology, each external source of financing is regressed on age, gender, educational level, fear of failing, know an
entrepreneur, investor protection, extractive sector, export intensity, expected number of jobs in 5 years, new product/few competitors, new technology,
improvement-driven opportunity, economic development level, and geographic region via a logistic model. (2) Under the propensity score matching methodology,
socially-oriented is regressed on the same covariates via a logistic model.

Appendix A. Appendix

A.1. Appendix 1A. GEM countries by geographic region and economic development level

Region Factor driven Efficiency driven Innovation driven

Botswana Egypt
Africa Burkina Faso Morocco
Cameroon South Africa
Senegal Tunisia
India China Australia
Iran Indonesia Israel
Asia & Oceania Philippines Kazakhstan Japan
Vietnam Lebanon South Korea
Malaysia Taiwan
Thailand
Bulgaria Belgium
Croatia Estonia
Hungary Finland
Latvia Germany
Macedonia Greece
Poland Ireland
Europe Romania Italy
Slovakia Luxembourg
Turkey Netherlands
Norway
Portugal
Slovenia
Spain
Sweden
Switzerland
UK
Argentina Puerto Rico
Barbados
Brazil
Chile
Latin America & Caribbean Colombia
Ecuador
Guatemala
Mexico
(continued on next page)

14
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

(continued )
Region Factor driven Efficiency driven Innovation driven

Panama
Peru
Uruguay
North America Canada
USA
Note: factor-driven economies are the least developed, and are dominated by subsistence agriculture and extraction businesses,
with a heavy reliance on unskilled labor and natural resources. Efficiency-driven economies in turn are increasingly competitive,
with more-efficient production processes and increased product quality; while innovation-driven economies are the most
developed, where businesses are more knowledge-intensive and the service sector expands.

A.2. Appendix 1B. Sampled countries by legal origin

Country Obs. % own money N◦ sources IP FD PR Trust Family Friends PI/VC Govt.

English
Australia 162 75.9 0.93 5.70 0.91 5.51 54.4 19.0 6.5 18.1 14.1
Barbados 234 81.4 0.65 3.00 0.47 5.02 – 25.6 7.0 5.3 3.7
Botswana 568 62.9 1.33 6.00 0.27 4.90 – 37.4 14.0 16.0 36.5
Canada 243 70.9 1.29 8.70 0.89 5.99 41.1 21.0 10.2 18.8 31.4
India 272 69.7 1.71 6.30 0.42 4.06 33.0 75.5 39.9 6.7 6.4
Ireland 131 71.4 1.36 8.30 0.67 5.87 – 18.5 6.4 25.3 38.8
Israel 102 61.8 1.06 8.30 0.60 4.81 23.0 35.2 9.1 10.4 10.3
Malaysia 44 64.8 1.07 8.70 0.66 5.30 8.5 45.7 2.2 2.2 19.6
Puerto Rico 111 85.1 0.70 7.00 – 5.92 22.4 16.5 3.6 4.8 18.4
South Africa 170 64.3 1.03 8.00 0.61 5.60 23.5 31.4 9.1 11.5 20.5
Thailand 294 85.1 1.28 7.70 0.72 4.09 32.6 69.2 13.9 6.7 3.6
United Kingdom 321 83.7 0.84 8.00 0.88 6.22 30.0 20.2 2.2 10.9 22.0
United States 224 73.4 1.38 8.30 0.90 5.35 38.2 23.5 14.2 22.8 21.8
Obs./Mean 2876 73.1 1.13 7.23 0.67 5.3 30.7 33.7 10.6 12.3 19.0

German
Bulgaria 33 79.4 1.06 6.00 0.37 3.47 19.4 60.3 20.6 16.9 0.0
China 295 91.4 1.80 5.00 0.63 4.49 62.7 67.4 31.7 25.7 14.9
Croatia 101 77.9 1.43 3.30 0.48 3.79 – 33.3 13.5 32.0 14.2
Estonia 144 64.5 1.39 5.70 0.28 5.18 38.5 28.1 12.1 27.2 20.6
Germany 121 72.1 0.94 5.00 0.74 5.64 42.0 16.8 6.4 28.1 12.3
Hungary 80 74.3 1.06 4.30 0.43 3.73 28.1 26.8 7.1 16.8 8.8
Latvia 154 76.1 0.92 5.70 0.25 4.59 23.9 35.7 8.1 20.8 8.6
Poland 105 71.2 1.32 6.00 0.48 4.32 22.3 23.7 4.5 28.6 13.5
Slovakia 117 71.4 1.33 4.70 0.32 3.84 25.8 25.5 12.1 32.5 19.0
Slovenia 42 72.7 1.17 7.30 0.38 4.18 20.0 47.2 4.2 18.6 14.1
S. Korea 142 76.1 1.34 4.04 0.84 4.21 29.7 38.3 8.8 45.0 12.2
Switzerland 70 65.9 1.14 3.00 0.98 6.24 49.4 28.4 13.6 28.2 16.5
Taiwan 110 65.3 0.71 6.30 0.63 5.73 29.6 14.0 19.4 12.4 11.6
Obs./Mean 1514 72.4 1.14 4.98 0.41 4.4 25.9 37.8 13.3 25.8 11.3

Scandinavian
Finland 80 70.1 1.43 5.70 0.75 6.39 58.0 24.7 7.2 15.1 31.2
Norway 69 72.8 0.46 6.70 0.67 5.99 73.7 3.5 1.2 10.6 14.1
Sweden 121 62.6 1.06 6.30 0.79 5.69 63.8 23.6 6.2 16.1 17.6
Obs./Mean 270 68.5 0.98 6.23 0.74 6.0 65.2 23.6 6.2 13.9 21.0

French
Argentina 165 80.5 0.72 5.00 0.30 2.62 22.7 28.1 4.7 7.7 9.8
Belgium 78 69.1 1.13 7.00 0.67 5.45 – 13.7 4.8 15.5 23.8
Brazil 203 90.0 0.48 5.30 0.63 3.99 6.5 22.3 4.5 2.1 1.6
Burkina Faso 680 46.9 0.82 3.70 0.12 3.50 14.4 31.0 6.2 5.6 20.7
Cameroon 480 69.9 0.81 4.30 0.10 3.64 – 34.6 21.8 2.4 7.3
Chile 1119 75.6 1.22 6.30 0.46 4.98 12.6 24.7 6.8 14.1 28.3
Colombia 728 59.4 1.70 8.30 0.38 3.86 4.1 35.7 12.1 17.9 24.9
Ecuador 583 74.3 0.91 – 0.18 – 7.2 30.2 3.5 3.4 3.7
Egypt 129 61.5 1.25 3.70 0.28 3.56 20.6 40.5 17.2 11.3 7.1
Greece 62 75.5 1.64 5.30 0.56 3.91 – 34.6 11.5 19.5 46.1
Guatemala 337 70.3 0.87 3.30 0.23 3.85 15.6 24.4 7.1 5.8 0.8
Indonesia 719 97.7 1.11 6.00 0.36 4.29 41.1 53.3 10.1 5.8 17.9
Iran 226 61.6 1.28 3.70 0.36 3.86 10.5 52.0 17.0 4.7 4.0
Italy 37 65.8 1.13 6.00 0.78 4.01 28.3 28.6 8.1 14.5 22.6
Kazakhstan 65 67.7 1.49 6.70 0.33 4.11 38.8 58.6 15.9 8.7 21.4
Lebanon 603 67.6 1.02 5.00 0.30 3.53 9.8 61.2 11.2 5.1 0.9
Luxembourg 88 64.8 1.23 4.30 0.74 6.11 – 26.4 7.7 15.1 29.3
Macedonia 62 68.0 1.44 7.00 0.33 4.55 13.3 39.3 15.9 22.4 20.8
Mexico 594 71.5 0.93 5.70 0.39 3.96 12.4 49.9 6.4 4.5 8.1
Morocco 62 63.2 0.91 4.70 0.35 4.85 12.5 44.4 7.3 6.1 3.7
Netherlands 90 70.5 0.66 4.70 0.71 5.84 66.2 15.3 4.6 8.5 10.9
Panama 198 93.2 0.69 5.30 0.33 4.50 – 45.0 12.4 0.0 1.2
(continued on next page)

15
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

(continued )
Country Obs. % own money N◦ sources IP FD PR Trust Family Friends PI/VC Govt.

Peru 350 72.4 1.00 7.00 0.37 3.54 8.2 43.0 5.3 2.5 1.0
Philippines 351 77.8 1.37 4.30 0.37 4.28 2.8 66.0 26.6 17.5 15.0
Portugal 105 72.4 0.94 6.00 0.69 4.85 – 29.7 9.5 10.3 16.7
Romania 128 66.2 1.29 6.00 0.31 3.96 7.1 37.4 9.0 14.9 26.5
Senegal 664 47.3 1.07 3.00 0.10 3.95 – 45.2 37.1 18.5 16.1
Spain 546 74.4 0.83 5.00 0.89 4.32 19.0 19.5 3.7 5.2 14.3
Tunisia 134 54.9 1.87 6.00 0.24 4.00 15.5 68.3 31.7 15.0 19.9
Uruguay 143 76.8 0.99 5.00 0.23 4.90 14.4 23.4 10.9 9.7 9.6
Vietnam 268 95.4 1.41 3.30 0.38 3.61 51.0 81.5 28.1 5.9 8.9
Obs./Mean 9997 71.0 1.10 5.23 0.40 4.2 18.9 39.0 12.2 7.7 9.8
Joint eq. F-test 13.7 19.4 7.9 7.0 9.91 10.6 22.9 11.1 28.5 42.2
P-value 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Notes (1) % own money and N◦ sources correspond with country average figures of TEA-own money provided or expected to be provided to the new business
(percentage) and TEA-No of external sources of financing; IP, FD, PR, Trust, with Investor protection (0–10 scale), Financial Development (0–1 scale), property rights
(1–7 scale) and with level of social trust attitude (%), respectively; and, Family, Friends, PI/V, and Govt. with country average figures of TEA_FRMa, TEA_FRMb,
TEA_FRMe, and TEA_FRMf, respectively (Appendix 2) (2) Null hypothesis under joint equality F-test is that means are equal across groups.

A.3. Appendix 1C. Ordered logistic model fitted to the number of external sources of financing

Marginal effects for selected covariates

Business angel
External sources dy/dx Std. Err z P-value
0 − 0.051 0.013 − 3.99 0.00
1–2 0.024 0.006 3.95 0.00
>2 0.027 0.007 3.98 0.00
Helped to start a business
External sources dy/dx Std. Err z P-value
0 − 0.122 0.018 − 6.70 0.00
1–2 0.058 0.009 6.51 0.00
>2 0.064 0.010 6.67 0.00
Home-based business
External sources dy/dx Std. Err z P-value
0 0.085 0.009 9.61 0.00
1–2 − 0.040 0.004 − 9.08 0.00
>2 − 0.045 0.005 − 9.50 0.00
Growth business
External sources dy/dx Std. Err z P-value
0 − 0.113 0.013 − 8.45 0.00
1–2 0.054 0.007 8.07 0.00
>2 0.059 0.007 8.38 0.00
New product/few competitors
External sources dy/dx Std. Err z P-value
0 − 0.041 0.009 − 4.53 0.00
1–2 0.019 0.004 4.47 0.00
>2 0.021 0.005 4.52 0.00
Factor-oriented economy
External sources dy/dx Std. Err z P-value
0 − 0.161 0.014 − 11.61 0.00
1–2 0.050 0.003 14.46 0.00
>2 0.111 0.013 8.84 0.00
Note: home-based business: no customers are located abroad; growth business: entrepreneur expects 20+ jobs in the
next 5 years.
N = 10,031 Prob >χ2 = 0.00 Pseudo R2 = 3.2%.

Appendix B. Appendix 2. GEM 2015 Global Adult Population Survey (APS) individual data, special topics

(a) Entrepreneurial finance

Early-stage business
Continuous variable
TEA_MONTOTUS How much money, in total, will be required to start this new business? (USD)
TEA_MONOWNUS_p Own money provided or expect to provide to the new business (percentage)
Binary variable =1 if yes; 0 otherwise
TEA_FRMa Family members (Received or expect to receive $ from)
TEA_FRMb Friends or neighbors (Received or expect to receive $ from)
TEA_FRMc Employer or work colleagues (Received or expect to receive $ from)
TEA_FRMd Banks or other financial institutions (Received or expect to receive $ from)
TEA_FRMe Private investors or venture capital (Received or expect to receive $ from)
TEA_FRMf Government programs, donations or grants (Received or expect to receive $ from)
TEA_FRMg Online crowd-funding (Received or expect to receive $ from)

Established business
(continued on next page)

16
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

(continued )
(a) Entrepreneurial finance

Continuous variable
EB_MONTOTUS How much money, in total, was required to start this new business? (USD)
EB_MONOWNUS_p Own money provided to the new business (percentage)
Binary variable =1 if yes; 0 otherwise
EB_FRMa Family members (Received $ from).
EB_FRMb Friends or neighbors (Received $ from)
EB_FRMc Employer or work colleagues (Received $ from)
EB_FRMd Banks or other financial institutions (Received $ from)
EB_FRMe Private investors or venture capital (Received $ from)
EB_FRMf Government programs, donations or grants (Received or $ from)
EB_FRMg Online crowd-funding (Received $ from)
(b) Social entrepreneurship

Sestart Are you, alone or with others, currently trying to start or currently leading any kind of activity that has a social,
environmental or community objective?
Scale 1 Yes, currently trying to start, 2 Yes, currently leading, 3 Yes, trying to start and leading, 4 No
Seecon For my organization generating value to society and the environment is more important than generating
financial value for the company.
Sesocial My organization puts more emphasis on social value than on environmental value.
Seinproc My organization offers a new way of producing a product or service.
Seprofit Profits will be reinvested to serve the social or environmental purpose of my organization.
Seimpact My organization puts substantial effort in measuring its social or environmental impact.
Scale 1 Strongly disagree, 2 Somewhat disagree, 3 Neither agree nor disagree, 4 Somewhat agree, 5 Strongly agree

References He, J., Nazari, M., Zhang, Y., & Cai, N. (2020). Opportunity-based entrepreneurship and
environmental quality of sustainable development: A resource and institutional
perspective. Journal of Cleaner Production, 256, 120390.
Amoros, J., Etchebarne, M., Zapata, I., & Felzensztein, C. (2016). International
Hoogendoorn, B., van der Zwan, P., & Thurik, R. (2020). Goal heterogeneity at start-up:
entrepreneurial firms in Chile: An exploratory profile. Journal of Business Research,
Are greener start-ups more innovative? Research Policy, 49(10), 104061.
69, 2052–2060.
Hörisch, J., Kollat, J., & Brieger, S. (2017). What influences environmental
Andersen, R. (2008). Modern methods for robust regression. Sage Publications, Inc.
entrepreneurship? A multilevel analysis of the determinants of entrepreneurs’
Bellavitis, C., Filatotchev, I., Kamuriwo, D., & Vanacker, T. (2017). Entrepreneurial
environmental orientation. Small Business Economics, 48, 47–69.
finance: New frontiers of research and practice. Venture Capital, 19, 1–16.
Kamal, S., & Daoud, Y. (2018). Explaining the gender gap in entrepreneurial propensity.
Bosma, N., Schøtt, T., Terjesen, S., & Kew, P. (2016). Global Entrepreneurship Monitor
In N. Faghi, & M. R. Zali (Eds.), Contributions to Management ScienceEntrepreneurship
2015/2016: Special Report on Social Entrepreneurship. Available at: http://www.
ecosystem in the Middle East and North Africa (MENA). Springer International
gemconsortium.org/report.
Publishing.
Bradač, B., & Crnogaj, K. (2020). Social impact, innovations, and market activity of social
Koellinger, P., Minniti, M., & Schade, C. (2013). Gender differences in entrepreneurial
enterprises: Comparison of European countries. Sustainability, 12(5), 1915. https://
propensity. Oxford Bulletin of Economics and Statistics, 75(2), 213–234.
doi.org/10.3390/su12051915
Korosteleva, J., & Mickiewicz, T. (2011). Start-up financing in the age of globalization.
Brown, R., & Rocha, A. (2020). Financing entrepreneurship in times of crisis: Exploring
Emerging Markets Finance and Trade, 47(3), 23–49.
the impact of COVID-19 on the market for entrepreneurial finance in the United
La Porta, R., Lopez-de-Silanes, F., & Schleifer, A. (2008). The economic consequences of
Kingdom. International Small Business Journal, 38, 380–390.
legal origins. Journal of Economic Literature, 46(2), 285–332.
Cameron, A., & Trivedi, P. (2005). Microeconometrics. Cambridge, U.K: Cambridge
La Porta, R., Lopez-de-Silanes, F., Schleifer, A., & Vishny, R. (1997). Legal determinants
University Press.
of external finance. Journal of Finance, 52(3), 1131–1150.
Cerulli, G., & Ventura, M. (2019). Estimation of pre- and post-treatment average
Langowitz, N., & Minniti, M. (2007). The entrepreneurial propensity of women.
treatment effects with binary time-varying treatment using Stata. Stata Journal, 19,
Entrepreneurship Theory and Practice, 31(3), 341–364.
551–565.
Lee, N., Sameen, H., & Cowling, M. (2015). Access to finance for innovative SMEs since
Cumming, D., Deloof, M., Manigart, S., & Wright, M. (2019). New directions in
the financial crisis. Research Policy, 44(2), 370–380.
entrepreneurial finance. Journal of Banking and Finance, 100, 252–260.
Leitch, C., Welter, F., & Henry, C. (2018). Women entrepreneurs’ financing revisited:
Cumming, D., & Groh, A. (2018). Entrepreneurial finance: Unifying themes and future
Taking stock and looking forward. Venture Capital, 20, 103–114.
directions. Journal of Corporate Finance, 50, 538–555.
Lepoutre, J., Justo, R., Terjesen, S., & Bosma, N. (2013). Designing a global standardized
Cumming, D., & Johan, S. (2017). The problems with and promise of entrepreneurial
methodology for measuring social entrepreneurship activity: The global
finance. Strategic Entrepreneurship Journal, 11, 357–370.
entrepreneurship monitor social entrepreneurship study. Small Business Economics,
Daniels, C., Herrington, M., & Kew, P. (2016). Global Entrepreneurship Monitor 2015/
40, 693–714.
2016: Special Report on Entrepreneurial Finance. Available at: http://www.gemco
Mickiewicz, T., & Rebmann, A. (2020). Entrepreneurship as trust. Foundations and Trends
nsortium.org/report.
in Entrepreneurship, 16(3), 244–309.
Denis, D. (2004). Entrepreneurial finance: An overview of the issues and evidence.
Mollick, E. (2014). The dynamics of crowdfunding: An exploratory study. Journal of
Journal of Corporate Finance, 10(2), 301–326.
Business Venturing, 29, 1–16.
Di Pietro, F., & Buttice, V. (2020). Institutional characteristics and the development of
Newbold, P., Carlson, W., & Thorne, B. (2020). Statistics for business and economics (9th
crowdfunding across countries. International Review of Financial Analysis, 71, 101543.
ed.). U.K: Pearson Education Limited.
Ding, Z., Au, K., & Chiang, F. (2015). Social trust and angel investors’ decisions: A
Nigam, N., Benetti, C., & Johan, S. (2020). Digital start-up access to venture capital
multilevel analysis across nations. Journal of Business Venturing, 30(2), 307–321.
financing: What signals quality? Emerging Markets Review, 45, 100743.
Estrin, S., Mickiewicz, T., & Stephan, U. (2016). Human capital in social and commercial
Nofsinger, J., & Wang, W. (2011). Determinants of start-up firm external financing
entrepreneurship. Journal of Business Venturing, 31, 449–467.
worldwide. Journal of Banking and Finance, 35, 2282–2294.
Fernandez-Laviada, A., Lopez-Gutierrez, C., & Perez, A. (2020). How does the
Oranburg, S. (2020). Start-up financing. In F. Pacheco-Torgal (Ed.) (2nd ed.,Woodhead
development of the social enterprise sector affect entrepreneurial behavior? An
Publishing Series in Civil and Structural EngineeringStart-up creation: The smart eco-
empirical analysis. Sustainability, 12, 826.
efficient built environment (pp. 59–79).
Fernandez-Laviada, A., Lopez-Gutierrez, C., & San Martin, P. (2020). The moderating
Parker, S. (2018). The economics of entrepreneurship (2nd ed.). U.K: Cambridge University
effect of countries’ development on the characterization of the social entrepreneur:
Press.
An empirical analysis with GEM data. Voluntas, 31, 563–580.
Pruthi, S. (2014). Social ties and venture creation by returnee entrepreneurs.
Frešer, B., & Tominc, P. (2018). Innovativeness and financial resources diversity of
International Business Review, 23(6), 1139–1152.
Slovenian early-stage entrepreneurs. Econviews, 31(1), 21–33.
Robb, A., & Watson, J. (2012). Gender differences in firm performance: Evidence from
Greene, W. (2018). Econometric analysis (8th ed.). Pearson Education.
new ventures in the United States. Journal of Business Venturing, 27, 544–558.
Hastings, D., & Anwar, N. (2019). Home-based businesses and the factors affecting their
Roper, S., & Scott, J. (2009). Perceived financial barriers and the start-up decision.
collaborative propensity. In Institute of Small Business and Entrepreneurship Conference
International Small Business Journal, 27, 149–171.
2019 ISBE.

17
V. Fernandez International Review of Financial Analysis 75 (2021) 101722

Shahriar, A. Z. M. (2018). Gender differences in entrepreneurial propensity: Evidence Vaznyte, E., & Andries, P. (2019). Entrepreneurial orientation and start-ups’ external
from matrilineal and patriarchal societies. Journal of Business Venturing, 33, 762–779. financing. Journal of Business Venturing, 34, 439–458.
Terjesen, S., Bosma, N., & Stam, E. (2016). Advancing public policy for high-growth,
female, and social entrepreneurs. Public Administration Review, 76(2), 230–239.

18