Access to finance for UK social enterprises

Abstract

We investigate access to finance for social enterprises, including those that are women and minority ethnic group (MEG) led. Using data from the UK Longitudinal Small Business Survey, we find that relative to small and medium enterprises (SMEs), social enterprises are less likely to apply for bank overdrafts, but more likely to apply for government grants. However, upon application, social enterprises are more likely to receive commercial mortgages, credit card funding, government grants and loans from mainstream financial intermediaries. By leadership diversity, women-led social enterprises are more likely to apply for loans from a bank, but less likely to receive bank funding compared to male-led counterparts. Our results also show that MEG-led social enterprises are less likely to apply for credit cards and government grants. Nevertheless, when they do apply, MEG-led enterprises have a higher probability of being granted government funding. In contrast, upon application, their chances of securing a bank overdraft facility are lower compared to those led by non-minority ethnic groups.

KEYWORDS:

• Access to finance
• sources of funding
• small- and medium-sized enterprises
• social enterprises
• women-led social enterprises
• minority-ethnic group enterprises

1. Introduction

Prior evidence suggests that small- and medium-sized enterprises (SMEs hereafter) face challenges in accessing sufficient external finance to fund day-to-day operations and longer-term strategic goals (Beck and Demirguc-Kunt 2006).¹ Such challenges can be more onerous and costly for ethnic minority and women-led SMEs (Blanchflower, Levine, and Zimmerman 2003; Coleman and Robb 2009; Fairlie, Robb, and Robinson 2021; Guzman and Kacperczyk 2019; Mascia and Rossi 2017) with negative implications for subsequent firm investment and growth (Brown et al. 2022). In this paper, we augment and complement prior research by investigating the underlying factors affecting access to finance for social enterprises, and whether this differs for those that have women or Minority Ethnic Group (MEG) leadership.²

Social enterprises generate income from trading activities, and use resulting profits to further social, ethical and environmental goals (Smith, Gonin, and Besharov 2013). Given the intersection of their respective commercial activities with ongoing social and environmental societal challenges, social enterprises have attracted the interest of academics and policymakers (Haugh et al. 2022; Hota 2023; Wilson and Post 2013). At an aggregate level, social enterprises augment and complement existing commercial for-profit and public sector goods and services provision by engaging in commercial activities that contribute to tackling issues related to aging, health, environment, and various forms of exclusion. They do this while generating profits to ensure financial viability.

Despite forming an important part of the SME ecosystem and promoting inclusive growth via employment creation, skills development and investment in local communities, current knowledge regarding social enterprises remains limited (Belz and Binder 2017; York, O'Neil, and Sarasvathy 2016). Social enterprises are involved in complex relationships with multiple stakeholders (customers, competitors, employees, funders, government, recipients, suppliers) emanating from diverse backgrounds (Austin, Stevenson, and Wei-Skillern 2006; Lumpkin and Bacq 2019). In common with mainstream SMEs, social enterprises face a myriad of challenges, including access to sufficient finance to fund day-to-day operations and longer-term strategic goals. However, relative to mainstream SMEs, financing constraints are likely to be more pronounced for social enterprises, given that their social mandate combined with a need to generate profit presents additional challenges to demonstrating creditworthiness to prospective lenders (European Parliament 2016). The ability of social enterprises to access external finance is an issue (Doherty, Haugh, and Lyon 2014), which has become even more pronounced in the wake of the: COVID-19 pandemic; the cessation of public policy business support measures introduced during the pandemic; and the ongoing cost of living crisis. Consequently, there is an urgent need for further research on access to finance for social enterprises, especially those that are MEG- and women-led (Carter et al. 2015; Lyon and Owen 2019).

As a setting for the current study, we use the United Kingdom, where the social enterprise sector accounts for approximately 3% of UK GDP, and is one of the fastest-growing forms of business, with over 100,000 organisations contributing £60 billion to the economy and employing over two million individuals (Social Enterprise UK 2018).³ In the UK, social enterprises are a core part of the wider SME population, conducting a variety of commercial activities across economic sectors and contributing to job creation (Haugh et al. 2022). Social enterprises are also characterised by a more diverse leadership than mainstream SMEs. According to Social Enterprise UK (2019), 40% of social enterprises are led by women, over twice that of mainstream SMEs (17%). Moreover, 13% of social enterprises are MEG-led, which also represents a much higher proportion relative to mainstream SMEs (5%). Given the general importance of social enterprises and specific importance of women- and MEG-led social enterprises for the UK economy, there are obvious economic and social development grounds for undertaking research on factors affecting access and use of various forms of finance, and whether this differs for those that are MEG- and women-led (Di Domenico, Haugh, and Tracey 2010; Doherty, Haugh, and Lyon 2014; Lee and Cowling 2013).

The data set used in the current study is the 2016–2019 Longitudinal Small Business Survey (LSBS) commissioned and published by the Department of Business, Energy and Industrial Strategy (Department for Business Energy and Industrial Strategy 2022). The LSBS is a large-scale cross-sectional and longitudinal telephone survey of owners/proprietors, managing directors or other senior directors in UK-based SMEs. The main advantage of using the LSBS as an information source is that the sample of SMEs is representative of the population of 5.5 million UK SMEs. Moreover, the LSBS uses a consistent classification method to identify social enterprises, and thus overcomes definitional challenges prevalent in prior research. Our data set also allows us to identify gender and ethnicity-based leadership characteristics (i.e. social enterprises that are women and minority ethnic group (MEG) led), and thus investigate their impact on access to finance.

We examine how social enterprises mobilise financial resources compared to mainstream SMEs, with a specific focus on the role of leadership diversity.⁴ The key research questions are: How does the use of various forms of finance differ between social enterprises and mainstream SMEs? Within social enterprises, how does leadership diversity in terms of women-led and minority ethnic group (MEG)-led social enterprises impact the use of different forms of financing? And how do social enterprises and their leadership diversity influence both the demand for (applying for) finance and access to finance? By addressing these questions, this paper aims to provide evidence on: the financial challenges faced by social enterprises; provide comparisons between social enterprises and mainstream SMEs regarding the demand and access to finance; and examine the impact of leadership diversity at social enterprises on financing. Overall, the goal is to advance academic and empirical understanding of financial resource mobilisation and access to finance issues for the social enterprise sector.

Our investigation proceeds in two stages. In stage 1, we examine the use of various forms of finance by social enterprises. We utilise probit models to estimate (relative to mainstream SMEs) the usage of different sources of finance by social enterprises including: bank overdrafts; commercial mortgages; credit cards; equity finance; factoring/invoice discounting; government or local authority grants; leasing or hire purchase; loans from a bank, building society or other financial institution; loans from family/friends; loans from a peer-to-peer platform; and loans from business partner/director/owner. We also consider the potential impact of leadership diversity (in the form of women-led or MEG-led business) on the type of finance used. While informative, the results of our stage 1 analysis do little to disentangle the differences between applying and receiving (or not receiving) finance. Given that the likelihood of a social enterprise accessing finance is conditional upon applying for it, a sample selection adjustment is necessary. Consequently, in stage 2 of our analysis, we utilise a Heckman sample selection probit model to investigate how being a social enterprise and its leadership diversity influence the demand for (applying for) finance and access to finance. This provides insights to the extent to which social enterprises have access to the appropriate forms of funding necessary to achieve economic, ethical, social and environmental objectives.

We contribute to several strands of literature. First, we contribute to evidence regarding financial resource mobilisation for social enterprises. Prior evidence suggests that accessing finance is an important barrier to the success of social enterprises (European Commission 2015), given that these entities are not perceived as viable clients by mainstream financial intermediaries (Doherty, Haugh, and Lyon 2014). Therefore, it is crucial that social enterprises have adequate access to external financial resources for the pursuit of their respective social, ethical and environmental mission (Doherty, Haugh, and Lyon 2014). Previous research also suggests there are significant impediments in the form of informational asymmetries, limited collateral and unstable cash flows to SMEs seeking bank funding (Berger and Black 2019; Berger and Udell 1998; Cowling, Liu, and Ledger 2012; Norden and Wang 2024). These impediments are likely to be more severe among social enterprises given their less conventional business model, where social, ethical and environmental goals augment conventional financial targets as an integral component of business strategy. Our results suggest that relative to mainstream SMEs, social enterprises are less likely to use bank overdrafts, loans (from either mainstream financial institutions or business partner/director/owner), equity finance and leasing or hire purchase, but are more likely to rely on grant funding provided by government and local authorities and to a lesser extent factoring / invoice discounting. Our analysis also provides insights to the demand (funding applications) and access to finance (outcomes of funding applications) for the main sources of finance used by social enterprises. Compared to mainstream SMEs, social enterprises are less likely to apply for bank overdrafts, and more likely to apply to government grants. However, upon application, social enterprises are more likely to receive commercial mortgage, credit card funding, government grants and loans from mainstream financial intermediaries.

Second, we contribute to the literature on access to finance for female entrepreneurs. The findings of prior academic research and various government inquiries suggest that relative to male counterparts, women-led businesses find it difficult to access external financing to establish and scale-up enterprises (Marlow and Patton 2005; Roper and Scott 2009; Rose 2019; 2022). Our results suggest that women-led social enterprises are less likely to use equity finance and loans from business partners/directors/owners. Considering the importance of leadership diversity of social enterprises for access to finance, our results suggest that women-led social enterprises are more likely to apply for loans from a bank, but conditional upon application, less likely to receive funding compared to male-led social enterprises.

Third, we contribute to the literature on access to finance for ethnic minority-led enterprises. Prior US evidence suggests that MEG-led SMEs are more likely to be: refused credit (Cavalluzzo, Cavalluzzo, and Wolken 2002; Fairlie, Robb, and Robinson 2021); pay more for credit (Blanchflower, Levine, and Zimmerman 2003); and be discouraged from applying for credit (Fairlie, Robb, and Robinson 2021; Neville et al. 2018). Furthermore, Kickul, Terjesen, and Justo (2013) suggest that social entrepreneurs tend to operate in resource-scarce environments. The prevalence of financing constraints leads to an organisational size gap emerging between white- and MEG-led firms (Barkley and Schweitzer 2023; Brown et al. 2022; Fairlie, Robb, and Robinson 2021). There is evidence that UK BAME-run businesses face more difficulty in accessing finance compared to non-BAME counterparts Recent reports highlight structural barriers and systemic disadvantages that prevent BAME entrepreneurs from accessing finance (British Business Bank 2020; Department for Communities and Local Government 2013; Federation of Small Business 2020). Access to finance represents a major barrier to the business success of Black, Asian and Other Ethnic Minority entrepreneurs, and is cited as reason why 39% and 49%, respectively cease working on their business idea (British Business Bank 2020).

Our results suggest that MEG-led social enterprises rely less on commercial mortgages, factoring/invoice discounting, government grants and leasing or hire purchase forms of finance compared to non-MEG-led counterparts, but are more likely to use bank overdrafts, equity finance, loans from mainstream financial intermediaries (such as a bank, building society or other financial institution) or loans from a business partner/director/owner. Our results also show that MEG-led social enterprises are less likely to apply for credit cards and government grants. However, conditional upon application, MEG-led social enterprises are less likely to be granted a bank overdraft facility but exhibit the largest probability of securing funding from a government or local authority grants. This combination of a paucity of bank-based funding and reliance on grant funding could affect the longer-term sustainability of MEG-led social enterprises.

Finally, we contribute to the growing literature on the importance of funding for firms with a social mandate (Austin, Stevenson, and Wei-Skillern 2006; Haugh et al. 2022; Lepoutre et al. 2013). Social entrepreneurs are often characterised by their ethic of care (André and Pache 2016), and assumed to be guided by ethical and moral considerations with the primary intention to help others (Pless 2012). Social enterprises have grown in prominence as they offer innovative solutions to pressing and complex social and environmental societal challenges (Lepoutre et al. 2013; Sarracino and Fumarco 2020; Zahra et al. 2009), while operating as commercial businesses and adding value to the economy via employment creation and investment. However, the likely trade-off between profit and purpose (social goals) faced by social enterprises may result in substantial financial resource constraints, which inhibit future tangible and intangible investments, employment creation and growth.

Overall, our findings have important implications for current and future policy towards social enterprises. Social enterprises represent a growing sector playing an important role in promoting the circular economy (OECD/European Commission 2022) and contributing to addressing the persistent social and environmental inequalities and the UK government levelling up agenda (Harrari and Ward 2022; UK Government 2022). Against this backdrop, the provision and access to appropriate forms of finance is crucial to ensuring that the financial sustainability and social mission of social enterprises is realised. In addition, our results suggest that the disadvantages faced by MEG- and women-led social enterprises in accessing finance could lead to this group of social enterprises failing to meet their full potential (Hyde 2021; Rose 2022).

The remainder of the paper is structured as follows. Section 2 provides a background on the evolution and policy towards UK social enterprises, and the importance of women- and MEG-leadership. Section 3 describes the data set and the research methodology. In section 4, we present the results of the empirical analysis. Section 5 provides a conclusion.

2. Background

2.1. Social enterprises

Social enterprises are, for the most part, SMEs engaged in the provision of goods and services with a wider social, ethical or environmental purpose.⁵ As such social enterprises play a vital role in stimulating entrepreneurial activity, increasing employment, building social capital and enhancing individual well-being, investing in disadvantaged areas, tackling social and financial exclusion, and addressing environmental and social challenges (Lepoutre et al. 2013; Sarracino and Fumarco 2020).

Social enterprises differ from traditional for-profit organisations, which utilise capital and labour inputs to produce goods and services with a primary aim of maximising profits. In contrast, social enterprises use labour and capital inputs to engage in entrepreneurial activity and produce goods and services in order to achieve social, ethical or environmental objectives that tackle problems related to poverty, deprivation, health and educational inequalities and environmental damage (Zahra et al. 2009). As such, social enterprises are a distinctive organisational form, which combine business activities with social, ethical and environmental goals.⁶

Social enterprises have formed an important part of the UK government policy agenda over the past 20 years. Teasdale (2012) provides an early discussion of the development of social enterprises in the UK. The scale and scope of social enterprises has increased in recent years in (part) response to gaps left in the provision of many goods and services following cuts to public services via government-imposed austerity programmes instituted in the aftermath of the global financial crisis. Social enterprises have emerged as a hybrid organisational form as the demarcations between the private, public, and non-profit sectors have eroded to become less distinct (Doherty, Haugh, and Lyon 2014). Overall, successive UK governments have undertaken a variety of measures to support the development and sustainability of social enterprises (Phillips 2006). More recently, the so-called Levelling Up agenda (Harrari and Ward 2022; UK Government 2022) presents an opportunity to inject more capital towards the social economy in the most left-behind communities.⁷

2.2. Debt finance, asset finance and alternative financing instruments

Social enterprises rely on multiple sources of finance to fulfil operational, cash flow and investment needs. There are several reasons why it is important for social enterprises to have access to different financing sources. Firstly, social enterprises often struggle to acquire external funding (Austin, Stevenson, and Wei-Skillern 2006; Lehner and Nicholls 2014; Schätzlein, Schlütter, and Hahn 2023), given that there are fewer financial institutions, instruments, and resources appropriately tailored to their needs. This can hinder their ability to effectively mobilise and utilise resources to achieve organisational objectives. Secondly, access to different financing sources is important for social enterprises because it funds growth, and signals commercial viability to profit-seeking investors (Reichert et al. 2021). Figure 1 offers definitions and descriptive summary of the finance instruments commonly used by social enterprises.

Figure 1. Forms of finance used by UK social enterprises.

Notes: This figure shows the various forms of debt finance, asset finance and alternative financing instruments typically used by Social Enterprises based on the LSBS survey. Cross-sectional survey weights from the LSBS survey have been applied to represent the population of SMEs in the UK. Respondents who answer ‘I do not know’ or ‘refused’ to answer are excluded from the sample. Traditional debt finance instruments: These represent the most common source of external finance for many SMEs, including social enterprises, as their use does not involve a sacrifice of ownership or control. The defining characteristic of these instruments is that they represent an unconditional claim on the borrower and should be repaid at an arranged later date, usually through regular repayments with added interest. (1) Bank loans: Bank loans are a relatively quick and straightforward way to secure the funding, with successful applications conditional upon overall creditworthiness and projected future performance. (2) Overdrafts: Bank overdrafts (or credit lines) and credit cards are a type of short-term flexible loan, up to an agreed limit provided by a financial intermediary. There is a fee payable with the use of any overdraft facility and interest paid on funds used. This represents an important source of funding for SMEs experiencing a temporary cash flow shortfall or requiring a cash boost because of short-term or unexpected situations. (3) Credit cards: Credit cards are an easily accessible and flexible source of funding for SMEs. They offer a source of revolving credit (within an agreed limit) for short-term business needs, and they are usually linked to a bank account. (4) Commercial mortgages: Commercial mortgage loans from high street banks (or specialist lenders) secured against commercial property or land for business purposes are also available to business owners. Asset-based finance instruments: Factoring and leasing are two types of asset-based finance instruments also available to social enterprises. These allow firms to obtain funding based upon the value of specific assets (such as trade accounts receivable, inventory, fixed assets, and real estate). As such, asset-based finance provides firms with access to cash (working capital) under flexible terms regardless of creditworthiness and projected future cash flows. The costs incurred are likely to be higher and the amount of funding received lower than that typically associated with conventional bank loans. (5) Factoring (invoice discounting): Factoring is a form of financing in which SMEs sell a certain amount of its receivables for an immediate payment at a discount. (6) Leasing: it is a form of financing that allows businesses to own fixed assets (e.g. real estate, vehicles, machinery & equipment) through leasing with a rental for a period of years until having full ownership at the end of the contract. Other sources of finance: it includes (7) Equity finance: Equity finance refers to all financial resources that are provided to firms (mainly growth-oriented and innovative start-ups) in return for an ownership interest. Family, friends, business angels and venture capitalists have been considered as the main providers of equity finance for SMEs. (8) Loans (from family, friends and related enterprises or owners): This unconventional form of business loan can provide funding at lower interest rates and fees, without the need to undergo onerous credit checks. (9) Crowdfunding/peer-to-peer lending (P2P): This type of finance has emerged as an alternative source of funding under which firms that are a member of an internet platform can borrow and lend money to one another directly, thus removing the need for a traditional financial intermediary. This type of funding option is typically unsecured, so is attractive to firms lacking collateral or credit history. (10) Grant funding: This is also a potential option for social and non-profit ventures. Given that funding is usually project-specific, and repayment is not always required, excessive reliance on this type of funding could potentially erode the financial self-sufficiency of these firms.

Figure 1 shows that bank loans, overdrafts and credit cards are used by approximately 6.4%, 23.5% and 24.7% of social enterprises.⁸ A small proportion of social enterprises rely on commercial mortgages (3.6%). Figure 1 also shows that 13% of social enterprises use leasing, while factoring (or invoice discounting) is used by approximately 3.3% of social enterprises. Equity finance refers to all financial resources that are provided to (mainly growth-oriented and innovative start-ups) firms in return for an ownership interest. Figure 1 suggests that a small proportion of (approximately 1.5%) of social enterprises rely on this type of finance. Loans from family, friends and related enterprises or owners are used by approximately 15.3% of social enterprises. Crowdfunding/peer-to-peer lending (P2P) is used by a mere 1.2% of social enterprises. Finally, approximately 7% of social enterprises use government or local authority grants as a source of finance.

2.3. Women-led and MEG-led businesses

In terms of access to finance for women-led social enterprises, the results presented later in this study have relevance for debates and evidence suggesting that female entrepreneurs face significant barriers to accessing finance (Azam Roomi, Harrison, and Beaumont-Kerridge 2009; Marlow and Patton 2005). Prior research suggests that relative to male counterparts, women-led firms are required to post higher levels of collateral and receive lower amounts of bank funding (Orhan 2001).

In the UK, the government-commissioned Rose Review finds that access to finance is a significant barrier to female entrepreneurship (Rose 2019; 2022). Women-led SMEs are established with significantly less capital than male-led counterparts. Moreover, female entrepreneurs are less aware of funding opportunities and are less likely to accrue significant debt. The Rose Review concludes that £250 billion of additional wealth could be added to the UK economy if women-led businesses were financed and grew at the same rate as male-founded enterprises. The results of prior academic research suggest that women-led businesses face higher costs of bank funding relative to male-led counterparts (Mascia and Rossi 2017). Consequently, many women-led enterprises rely significantly more on informal forms of funding (Coleman and Robb 2009).

Despite having less access to finance than white-led counterparts, MEG-led SMEs play a crucial role in adding value to the UK economy (Federation of Small Business 2020) via employment and innovation (British Business Bank 2020). Access to finance challenges have been found to arise from insufficient collateral, lack of credit history and language barriers (BDRC Continental 2017; Department for Communities and Local Government 2013). Prior research suggests that MEG-led firms are more likely to be refused credit. These firms are also more likely to be discouraged from applying for credit (Fraser 2009). However, more recent evidence suggests that immediately preceding and following the onset of the COVID-19 pandemic, ethnicity was not a significant factor in determining the success of loan applications (Cowling, Liu, and Conway 2023). Challenges to accessing finance facing women-led or MEG-led SMEs are even more pronounced for ethnic minority female entrepreneurs (Hyde 2021). These constraints are likely to limit the full potential of MEG-led SMEs to contribute to employment creation, capital accumulation and economic growth (British Business Bank 2020). The results presented in this study augment recent evidence, which suggests that traditional routes to gaining access to information, networks and finance do not recognise the needs of MEG-led social enterprises (Sepulveda and Rabbevåg 2021).

3. Data and methodology

3.1. Data

The UK Longitudinal Small Business Survey (LSBS) is the primary data source used in the present study. Commissioned by the Department for Business, Innovation and Skills (BEIS), the LSBS is a large-scale telephone survey of owner/proprietors, managing directors or other senior directors in UK-based SMEs.

The LSBS, initiated in 2015 by the UK Department for Business, Energy and Industrial Strategy (BEIS), is an annual survey targeting UK businesses with fewer than 250 employees. It classifies SMEs by number of employees, as follows: 1–9 for micro firms, 10–49 for small enterprises, and 50–249 for medium-sized enterprises. To ensure representativeness across regions, sectors, and size categories, the survey adds new respondents annually to offset panel depletion. The data is annually weighted cross-sectionally in order to provide a comprehensive picture of UK SMEs. The UK government relies on the Longitudinal Small Business Survey (LSBS) as a key source of data for understanding the scale and characteristics of social enterprises, and how they compare to the wider SME population (DCMS/BEIS 2017; 2021).

The LSBS database comprises both cross-sectional and panel data from annual surveys conducted from 2015 to 2020. Our analysis, however, focuses on the sample period from 2016 to 2019. This sample period selection is determined by two factors. Firstly, the 2015 Survey wave is intentionally excluded from the sample because of changes to the questionnaire after 2015, which do not allow us to draw comparisons over time. Secondly, the 2020 Survey wave is also excluded as it lacks specific information necessary for identifying social enterprises, which is available only in the 2017 and 2019 Survey waves (See Section 3.2. for a more detailed explanation). Our decision to include the 2016 and 2018 Survey waves in our sample, despite the introduction of social enterprise survey modules in 2017 and 2019, aligns with our need to take advantage of the survey's longitudinal nature. By incorporating these years, we can utilise lagged variables, thereby helping to address potential endogeneity concerns. Consequently, the sample period of 2016–2019 not only includes years when social enterprises are identifiable (2017 and 2019), but also their preceding years (2016 and 2018). The total number of observations of the raw LSBS database across the four years under analysis (2016 to 2019) is 41,884, broken down as follows: 9248 in 2016, 6619 in 2017, 15,015 in 2018, and 11,002 in 2019. Additionally, throughout these four years, 10,540 SMEs have participated in the Survey at least twice. The longitudinal aspect of the LSBS survey enables us to track social enterprises over time, across various UK regions and industry sectors.

3.2. LSBS classification of social enterprises

Prior estimates of the scale of UK social enterprises have been based largely upon results from the Small Business Survey (SBS), which was replaced by the LSBS in 2015 (Department for Business Energy and Industrial Strategy 2022). The most common definition of social enterprise used by the UK government is: ‘A social enterprise is a business with primarily social objectives whose surpluses are principally reinvested for that purpose in the business or in the community, rather than being driven by the need to maximise profit for shareholders and owners’ (Department of Trade and Industry 2002).

In 2017, the LSBS introduced a new module to identify businesses as social enterprises. Specific questions to identify social enterprises are included in the Survey every other year, and thus were included in the 2019, but not in the 2020 wave of the LSBS survey. The LSBS defines four types of organisations based on social and environmental goals, comprising: social enterprises; traditional non-profit enterprises; socially orientated SMEs; and commercial SMEs.⁹

The identification of social enterprises is based upon four key characteristics, comprising: income generated from trading; charitable status & legal form; use of surpluses/profits; and organisational goals (social/environmental/financial). Based on the LSBS classification (see Figure 2), social enterprises are classified as enterprises that have identifiable social/environmental goals; generate income from trading activities (i.e. engage in entrepreneurial activity); and use surplus/profit to further social/environmental goals. Social enterprises also include organisations that pursue social goals and generate more than 50% of income from trading activities. Socially oriented SMEs (so-called profit-with-purpose’ businesses) are enterprises that have social/environmental goals and generate income chiefly from trading activities, but do not use profits to further those goals. Traditional non-profits (which are mostly charities) are organisations that pursue social goals, but generate less than 50% of income from trading activities.¹⁰ Commercial SMEs have clear financial objectives and do not use profits to further social, ethical or environmental objectives.

Figure 2. Decision tree to identify social enterprises.

Notes: This figure summarises the decision process used by the LSBS to identify and classify social enterprises. ‘For-profit’ legal forms include sole proprietorship/trader, private limited company (by shares), public limited company, private unlimited company, foreign company. ‘Other’ legal forms include partnerships, limited liability partnerships, private company (limited by guarantee), co-operative, ‘other,’ do not know and refused answers. ‘Social’ legal forms include community interest company (limited by guarantee or shares), friendly society, industrial and provident society, trust, unincorporated association, community benefit society, charitable un/incorporated organisation. ‘Env.’ – Environmental. S/E – social or environmental. The figure is sourced from the Longitudinal Small Business Survey Year 3 (2017): Technical Report under the Open Government Licence v3.0. Source: BEIS (2018).

Figure 3, based on the full LSBS sample, shows that commercial SMEs represent around 70.18% of the business population in the UK, followed by socially oriented SMEs often referred to as ‘profit-with-purpose’ businesses (18.16%), social enterprises (8.1%) and traditional non-profits (3.5%).

Figure 3. UK SME ecosystem by organisational form.

Notes: In this study, we use a broad definition of SMEs which comprises socially oriented SMEs (so-called profit-with-purpose’ businesses) and commercial SMEs in line with the 2017 Social Enterprise Market Trends report published by the Department for Digital, Culture, Media and Sport (DCMS). The report is available at: https://www.gov.uk/government/publications/social-enterprise-market-trends-2017.

3.3. Descriptive statistics: SME characteristics and organisational forms

The LSBS encompasses detailed information on the characteristics of mainstream SMEs and social enterprises. A detailed definition of all the variables used in the empirical analysis is presented in Table 1.¹¹ A key dependent variable used in this study is a dummy variable that measures whether businesses in the sample are social enterprises or mainstream SMEs.

Table 1. Variable definition.

Variable	Definition	LSBS code
Classification of enterprises		SOCENT
Mainstream SME (base category)	This is a dummy variable that takes a value of 1 if the business is a Social Enterprise, and 0 if the business is a mainstream SME.
Social Enterprise
Women-led	Women-led businesses are defined as those majority-led by women, which is controlled by a single woman or having a management team of which a majority are women. ‘Majority’ here means over 50%.	WLED
Minority ethnic-led	A business where at least half of the leadership team comes from minority ethnic groups (as this is a UK survey, minority ethnic groups are those that are not White British, where White British includes White English, White Scottish etc.). The leadership team comprises the directors and working owners. We can include members of several ethnic groups and can include people who describe themselves as mixed ethnicity where White British is one of those ethnicities.	MLED
Aims to grow	Aim to grow sales over the next 3 years.	R1
Size		A2SPSS1
Zero employees (base category)	Zero employee business had no employees on their payroll (excluding owners and partners) at the time of the interview.
Micro	1–9 employees.
Small	10–49 employees.
Medium	50–249 employees.
Business age	Age of the firm.	A6SUM and A6, missing values for 2016 are completed with values from 2015
0–5 years (base category)
6–10 years
11–20 years
20+ years
Turnover change	Turnover in the past 12 months, compared with the previous 12 months.	P2
Decreased (base category)
Stayed roughly the same
Increased
Profit	Firm generates a profit or surplus after considering all sources of income in the last fiscal year.	P12
Urban area	Broad urban/rural categorisation from postcode.	URBRUR2
Family-owned	Business is a family-owned business (i.e. one which is majority-owned by members of the same family).	A12
Business plan	The business has a formal written business plan.	F5
Partnership
Region	Region where the firm has its headquarters.	NATION
England (base category)
Scotland
Wales
Northern Ireland
Sector	Industry Sector	SECTOR
Manufacturing sector (base category)	Production and construction (SIC 2007: ABCDEF).
Transportation and retail services	Transport, retail, and food service / accommodation (SIC 2007: GHI).
Business services	Business services (SIC 2007: JKLMN).
Other services	Other services (SIC 2007: PQRS).

Notes: This Table shows names and definitions of explanatory variables. All variables were gathered from the Longitudinal Small Business Survey, 2016–2019.

In this paper, commercial and socially oriented SMEs (so-called profit-with-purpose businesses or mission-led businesses) are combined in the joint category called ‘mainstream SMEs.’ This classification aligns with the methodology used in DCMS/BEIS (2017). The purpose of this classification is to facilitate a comparison between social enterprises and the wider SME sector. The justification for this classification lies in the core objectives of organisations and the manner in which profits are allocated. Commercial and socially oriented SMEs, while differing in their secondary goals, share a primary focus on financial sustainability and profit generation. In common with social enterprises, socially oriented SMEs may also pursue social or environmental goals. However, these do not predominantly guide their profit allocation. This similarity in profit orientation aligns them with commercial SMEs (rather than with social enterprises), whose primary objective is profitability. In contrast, social enterprises are characterised by their commitment to reinvesting a significant portion of their profits to advance social and environmental objectives. This commitment is integral to their identity and mission. Unlike commercial and socially oriented SMEs, social enterprises have business models inherently structured around social or environmental goals. Their operational focus is not just on financial viability, but also on making a substantial impact in their chosen societal or environmental areas. Thus, our categorisation recognises that while socially oriented SMEs may seek social or environmental impacts, this pursuit does not fundamentally change their operational focus or business model, which is closer to that pursued by commercial SMEs. In summary, the key distinction in our categorisation lies in the primacy of organisational objectives and the use of profits, which sets social enterprises apart from other SMEs.

Our estimable models defined in the following section include several control variables related to the demographic and managerial characteristics of the mainstream SMEs and social enterprises in the sample. Table 2 presents summary information. Mainstream SMEs (the benchmark category in our empirical analysis) represent 91.6% of our sample, while social enterprises represent 8.4% of the sample.¹² Table 2 presents a comparative analysis of mainstream SMEs and Social Enterprises across various metrics. In leadership diversity, Social enterprises show a slightly higher representation of women-led (20%) and minority ethnic-led (9.8%) organisations compared to mainstream SMEs (18.3% women-led, 4.5% minority ethnic-led). When it comes to control variables, a higher percentage of social enterprises aim for growth (60.1%) versus mainstream SMEs (51.6%). Size-wise, Social enterprises tend to be smaller, with a greater proportion having zero employees. They also exhibit a younger business age profile, with more enterprises in the ‘0–5 years’ category, and a higher likelihood of increased turnover and profitability than mainstream SMEs. Business Characteristics reveal that a slightly higher percentage of mainstream SMEs operate in urban areas and are family-owned, whereas social enterprises are more likely to have a business plan. Regionally and sector-wise, the distributions are similar for both types of enterprises, with notable differences in certain sectors like ‘Other services,’ where Social Enterprises have a higher representation. This analysis underscores the distinctive operational and strategic differences between social enterprises and mainstream SMEs. Table 3 presents correlations between the explanatory variables used in the empirical analysis. The highest pair-wise correlation is 0.34 (between size and business plan dummy), suggesting that multicollinearity issues are not a concern in the baseline model specification used in our empirical analysis.

Table 2. Summary statistics.

	Full Sample			Mainstream SMEs			Social Enterprises
	Mean	Std. Dev.	N	Mean	Std. Dev.	N	Mean	Std. Dev.	N
CLASSIFICATIONS OF ENTERPRISES
Mainstream SME (base category)	0.916	0.278	16,650
Social Enterprise	0.084	0.278	16,650
LEADERSHIP DIVERSITY
Women-led	0.202	0.402	38,479	0.183	0.387	14,322	0.200	0.400	1525
Minority ethnic-led	0.048	0.215	37,262	0.045	0.206	13,762	0.098	0.298	1436
CONTROL VARIABLES
Entrepreneur orientation
Aims to grow	0.531	0.499	40,984	0.516	0.500	14,725	0.601	0.490	1902
Size
Zero employees (base category)	0.759	0.428	40,984	0.766	0.423	14,725	0.713	0.453	1902
Micro (1–9)	0.198	0.399	40,984	0.194	0.395	14,725	0.220	0.414	1902
Small (10–49)	0.037	0.188	40,984	0.034	0.182	14,725	0.058	0.233	1902
Medium (50–249)	0.006	0.078	40,984	0.006	0.075	14,725	0.010	0.098	1902
Business age
0–5 years (base category)	0.176	0.381	40,842	0.144	0.351	14,685	0.101	0.302	1897
6–10 years	0.188	0.391	40,842	0.192	0.394	14,685	0.180	0.384	1897
11–20 years	0.268	0.443	40,842	0.292	0.455	14,685	0.281	0.450	1897
20+ years	0.369	0.482	40,842	0.373	0.484	14,685	0.438	0.496	1897
Turnover change
Decreased (base category)	0.225	0.417	38,992	0.236	0.425	14,116	0.181	0.386	1836
Stayed the same	0.489	0.500	38,992	0.481	0.500	14,116	0.474	0.499	1836
Increased	0.287	0.452	38,992	0.283	0.450	14,116	0.344	0.475	1836
Profitability
Profit	0.809	0.393	38,594	0.817	0.386	14,122	0.865	0.342	1868
Business characteristics
Urban area	0.697	0.459	40,934	0.693	0.461	14,692	0.674	0.469	1896
Family owned	0.876	0.330	40,797	0.893	0.310	14,670	0.688	0.463	1897
Business plan	0.295	0.456	39,603	0.270	0.444	14,243	0.410	0.492	1845
Region
England (base category)	0.882	0.323	40,984	0.884	0.321	14,725	0.879	0.327	1902
Scotland	0.058	0.234	40,984	0.058	0.234	14,725	0.059	0.235	1902
Wales	0.037	0.189	40,984	0.037	0.188	14,725	0.033	0.180	1902
Northern Ireland	0.023	0.149	40,984	0.022	0.145	14,725	0.029	0.168	1902
Sector
Manufacturing sector (base category)	0.260	0.439	40,984	0.264	0.441	14,725	0.220	0.415	1902
Transportation and retail services	0.191	0.393	40,984	0.194	0.395	14,725	0.181	0.385	1902
Business services	0.336	0.472	40,984	0.354	0.478	14,725	0.227	0.419	1902
Other services	0.213	0.410	40,984	0.188	0.390	14,725	0.372	0.483	1902

Notes: This table reports the summary statistics using data from the Longitudinal Small Business Survey, 2016–2019. Cross-sectional survey weights applied to represent the population of SMEs in the UK. The category mainstream SME includes both commercial SMEs and socially oriented SMEs. Traditional non-profit SMEs (which are mostly charities) and respondents who answer ‘I do not know’ or ‘refused’ to answer are excluded from the sample and subsequent empirical analysis. Variable definitions are reported in Table 1.

Table 3. Correlation matrix.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)
(1) SME type	1.000
(2) Women-led	0.049*	1.000
(3) MEG-led	0.024*	−0.006	1.000
(4) Aim to grow	0.029*	−0.030*	0.031*	1.000
(5) Firm size	0.099*	−0.029*	0.014*	0.276*	1.000
(6) Firm Age	0.049*	−0.052*	−0.067*	−0.074*	0.205*	1.000
(7) Turnover change	0.037*	−0.012	−0.016*	0.183*	0.158*	−0.088*	1.000
(8) Profit	−0.011	−0.038*	−0.031*	0.019*	0.055*	0.035*	0.186*	1.000
(9) Urban	0.000	0.007	0.098*	0.060*	0.089*	−0.042*	−0.004	−0.021*	1.000
(10) Family business	−0.201*	0.064*	−0.005	−0.114*	−0.298*	−0.072*	−0.068*	0.062*	−0.128*	1.000
(11) Business plan	0.128*	0.013	0.022*	0.222*	0.343*	−0.009	0.111*	−0.030*	0.066*	−0.222*	1.000
(12) Legal Status	−0.118*	−0.102*	0.012	0.120*	0.175*	0.014*	0.048*	0.074*	−0.045*	−0.007	0.085*	1.000
(13) Region	0.016	−0.012	−0.055*	0.000	−0.015*	−0.034*	0.011	0.000	−0.112*	0.035*	−0.011	−0.029*	1.000
(14) Broad Sector	0.166*	0.170*	0.058*	−0.022*	−0.005	−0.062*	0.001	−0.048*	0.149*	−0.169*	0.118*	−0.124*	−0.068*	1.000

Note: This table reports the correlation matrix between all variables used in this study.

*p < 0.01.

3.4. Empirical methodology

The present study utilises four recent waves (2016–2019) of the LSBS. We exploit the longitudinal element of the survey, and thus deal with endogeneity concerns by using lagged variables in our regression analysis. In order to investigate the access to and usage of various forms of finance by social enterprises, we rely on probit and sample selection (Heckman) probit models.

3.4.1 Types of finance used by social enterprises

In Stage 1 of our empirical analysis, probit models are used to investigate the determinants of the current use of different financing sources of mainstream SMEs and social enterprises. Here, the dependent variable is equal to one if a business i is using a specific source of finance (i.e., Bank Overdraft, Commercial Mortgage, Credit Cards, Equity Finance, Factoring/Invoice Discounting, Government or Local Authority Grants, Leasing or Hire Purchase, Loan from a Bank, Building Society or Other Financial Institution, Loan from Family/Friend, Loan from a Peer-to-Peer Platform, or Loan from Business Partner/Director/Owner) at time t (survey wave 2016-2020), and zero otherwise. (1) $\begin{array}{r}Pr(\mathrm{Finance}\mathrm{\_Sourc}{e}_{\mathrm{it}}=1)=\mathrm{\Phi }(X_{\mathrm{it}}\beta +v_{\mathrm{it}})\end{array}$(1) ${\nu }_{\mathrm{it}}$ are i.i.d., $N(0,{\sigma }_{\nu }^2)$, and $\mathrm{\Phi }$ is the standard normal cumulative distribution function. We include a wide range of covariates that prior theory suggests are likely to affect the decision to use various sources of finance by social enterprises. These include firm size, age, along with various other firm-level characteristics (such as women- or MEG- leadership), year, industry and regional fixed effects.¹³ In addition, our empirical approach uses lagged independent variables to mitigate endogeneity concerns arising from reverse causality for growth ambition, changes in turnover, profitability, and management characteristics (women-led and minority ethnic-led SMEs). All results associated with these models are presented in terms of average marginal effects (AME) and errors are clustered at regional level to account for correlations of any unobserved components of outcomes of SMEs located within the same cluster or geographical area.¹⁴

3.4.2. Access to finance

In stage 2 of our analysis, we investigate the drivers of funding applications and their resultant outcome using a probit model with sample selection (Van de Ven and Van Praag 1981), which applies the Heckman (1979) approach to correct for selection bias in probit-type equations. The basic indicator of access to finance is whether SME j (either mainstream SME or social enterprise) is successful in the applications for finance. However, given that the likelihood of a firm receiving funding is conditional upon applying for it, a sample selection adjustment is necessary. As a consequence, we follow previous literature and use a Heckman correction for selection (Lee and Brown 2016; Lee and Drever 2014). This assumes that there is an underlying relationship (latent equation) (2) $\begin{array}{r}{\text{Y}}_{\text{jt}}^{\ast }={\text{X}}_{\text{jt}}\beta +{\mu }_{1\text{jt}}\end{array}$(2) such that we observe only the binary outcome (outcome equation: successful finance application by SME j in time t). (3) $\begin{array}{r}{\text{y}}_{\text{jt}}^{\text{probit}}=({\text{y}}_{\text{jt}}^{\ast }>0)\end{array}$(3) The dependent variable, however, is not always observed. Rather, the dependent variable for SME j is observed if (selection equation: SME j applied for finance) (4) $\begin{array}{r}{\text{Y}}_{\text{jt}}^{\text{select}}=({\text{Z}}_{\text{jt}}\gamma +{\mu }_{2\text{jt}}>0)\end{array}$(4) where ${\mu }_1\sim \text{N}(0,1)$; ${\mu }_2\sim \text{N}(0,1)$; $\text{corr}({\mu }_{1,}{\mu }_2)=\rho (\text{rho})$. The model requires that the vector $(Z_{\mathrm{jt}})$ in the selection equation (which estimates the probability of applying for finance) contains an instrumental variable which should be excluded from the outcome regression (which estimates likelihood of obtaining finance, corrected for the likelihood of applying). Seeking any form of business advice by the SME in the last 12 months was used as selection variable.

We include a wide range of independent variables $(X_{\mathrm{jt}})$ and $(Z_{\mathrm{jt}})$, which are expected to affect the decision to use various sources of finance by social enterprises. These include firm size, age, along with various other firm-level characteristics (such as women- or MEG- leadership), year, industry and regional fixed effects. All results associated with these models are presented in terms of average marginal effects (AME) and errors are clustered at regional level to account for correlations of any unobserved components of outcomes of SMEs located within the same cluster or geographic area.

4. Results

In this section, we present the main results derived from the analysis of the LSBS (section 4.1).¹⁵ Next, we present the empirical results associated with the differential effect of social enterprises and the diversity of their leadership in using specific types of finance sources, but also the impact on access to finance (section 4.2).

4.1. Use of different types of finance by social enterprises

We commence by comparing social enterprises to mainstream SMEs with respect to the use of various forms of finance including bank overdrafts, commercial mortgages, credit cards, equity finance, factoring/invoice discounting, government or local authority grants, leasing or hire purchase, loans from a bank, building society or other financial institution, loans from family/friends, loans from a peer-to-peer platform, and loans from business partner/director/owner. Then, we focus our analysis on the influence of leadership diversity (women-led and MEG-led) within the sample of social enterprises on their use of various forms of finance.

4.1.1. Finance use by social enterprises versus mainstream SMEs

The results presented in Table 4 suggest that compared to mainstream SMEs, social enterprises are 2.8% less likely to rely upon bank overdrafts, and 5.6% less likely to use leasing or hire purchase. Social enterprises rely more on factoring and invoice discounting and less on equity finance relative to mainstream SMEs, albeit the differential impact in economic terms is small (0.7% and 0.8%, respectively). The positive effect on factoring and invoice discounting funding is well aligned with the needs of social enterprises given that it allows them to obtain finance based on the value of accounts receivables rather than relying on an externally generated credit rating. Our results also suggest that social enterprises have a 7.4% higher probability of using grants as a funding source compared to mainstream SMEs.

Table 4. Marginal effects of social enterprises with respect to mainstream SMEs on use of finance.

	Bank overdraft	Commercial Mortgage	Credit Cards	Equity Finance	Factoring/Invoice discounting	Government or local authority grants	Leasing or hire purchase
Social enterprise	−0.028***	−0.002	−0.013	−0.008*	0.007**	0.074***	−0.056***
	(−3.95)	(−0.47)	(−0.81)	(−1.73)	(2.48)	(8.64)	(−10.62)
Women-ledt–1	−0.030***	0.000	−0.009	−0.011***	0.001	0.002	−0.020*
	(−2.70)	(0.03)	(−1.22)	(−3.21)	(0.66)	(0.76)	(−1.75)
Minority ethnic-ledt–1	−0.008	0.032***	−0.031**	0.008***	−0.005**	0.002	−0.019
	(−0.58)	(6.62)	(−2.01)	(4.26)	(−2.34)	(0.13)	(−1.32)
Aims to growt–1	0.037***	−0.006	0.059***	0.013***	0.029***	0.022***	0.058***
	(3.83)	(−0.70)	(11.26)	(8.17)	(3.66)	(6.65)	(7.29)
Size: Micro	0.077***	0.035***	0.045***	−0.005	0.029***	0.025***	0.104***
	(8.51)	(6.62)	(16.24)	(−1.38)	(7.93)	(20.76)	(17.18)
Size: Small	0.065***	0.080***	0.120***	0.005	0.075***	0.037***	0.254***
	(11.37)	(6.45)	(28.80)	(1.38)	(14.24)	(5.41)	(63.25)
Size: Medium	0.079***	0.113***	0.181***	0.021***	0.116***	0.034***	0.303***
	(9.55)	(12.39)	(16.31)	(3.73)	(11.68)	(4.02)	(26.43)
Business age: 6–10 years	0.027	0.023***	0.067***	−0.013***	0.024**	0.011	0.023***
	(1.47)	(2.89)	(5.83)	(−3.55)	(2.14)	(0.89)	(6.34)
Business age: 11–20 years	0.087***	0.048***	0.141***	−0.008***	0.004	0.015	0.044***
	(5.45)	(5.58)	(6.73)	(−3.23)	(0.35)	(1.61)	(5.26)
Business age: 20+ years	0.099***	0.046***	0.137***	−0.025***	−0.004	0.025***	0.053***
	(6.68)	(7.68)	(24.60)	(−13.43)	(−0.28)	(2.72)	(6.14)
Turnover change (stayed the same)t–1	−0.034***	−0.009*	−0.013*	0.001	−0.010***	−0.011*	0.012**
	(−4.14)	(−1.79)	(−1.86)	(0.46)	(−4.11)	(−1.93)	(2.48)
Turnover change (increased)t–1	−0.015**	0.006	−0.001	0.008**	−0.005	0.006	0.014
	(−2.53)	(1.03)	(−0.03)	(2.33)	(−1.43)	(1.10)	(1.50)
Profitt–1	−0.064***	0.009	−0.016***	−0.018***	−0.008***	−0.025***	−0.009
	(−4.88)	(1.23)	(−2.81)	(−4.45)	(−2.81)	(−4.54)	(−0.82)
Locationt: Urban area	−0.021*	−0.011***	−0.022**	−0.001	0.009***	−0.026***	−0.042***
	(−1.78)	(−2.67)	(−2.04)	(−0.27)	(2.75)	(−13.19)	(−11.03)
Family owned	0.075***	0.024***	−0.018	−0.025***	0.010***	−0.019***	−0.002
	(6.98)	(19.20)	(−1.56)	(−10.41)	(3.08)	(−3.40)	(−0.14)
Business plan	0.041***	0.019***	0.035***	0.013***	0.020***	0.029***	0.017***
	(10.07)	(7.55)	(9.15)	(7.17)	(2.72)	(3.95)	(4.05)
Fixed effects
Year / Regional / Industry FEs	YES	YES	YES	YES	YES	YES	YES
Regional*Year / Industry*Year	YES	YES	YES	YES	YES	YES	YES
N	9525	9525	9525	9525	9525	9525	9525
Log pseudo-likelihood	−5733.453	−2422.589	−6098.964	−1040.055	−1968.478	−1676.575	−4916.598
R^2	0.690	0.921	0.641	0.974	0.939	0.948	0.747
AIC	11,472.905	4851.179	12,203.929	2086.111	3942.956	3359.150	9839.195
BIC	11,494.390	4872.664	12,225.414	2107.596	3964.441	3380.635	9860.680
Non-event (failures = 0)	6556	8769	5906	9273	8943	9027	7039
Event (successes = 1)	2969	756	3619	252	582	498	2486

Notes: This table shows average marginal effects (AMEs) from a probit model of SMEs characteristics on the probability of using various sources of debt. All regressions include a constant term. The base categories for categorical variables are: zero employees (size), 0–5 years (business age), 18–30 years old (owner’s age), decreased (turnover change). All models include year, industry, regional, industry*year and regional*year fixed effects. Z-statistics adjusted for clustering at regional level are reported in parentheses. Statistical significance at the 10%, 5%, and 1% levels are showed by *, ** and ***.

Table 5 presents results for various categories of loans. The most important finding is that compared to mainstream SMEs, social enterprises are 3.4% less likely to use loans from mainstream financial intermediaries, and 5.4% less likely to use loans from business partners, directors or owners.

Table 5. Marginal effects of social enterprises with respect to mainstream SMEs on use of finance.

	Loan from a bank, building society or other financial institution	Loan from family/friend	Loan from a peer-to-peer platform	Loan from business partner/director/owner
Social enterprise	−0.034**	−0.002	−0.004	−0.054***
	(−2.11)	(−0.30)	(−0.82)	(−8.56)
Women-ledt–1	−0.024***	0.015***	−0.009***	−0.015**
	(−3.75)	(2.98)	(−2.81)	(−2.33)
Minority ethnic-ledt–1	0.019***	0.048***	0.011***	0.025**
	(3.68)	(10.32)	(8.01)	(2.08)
Aims to growt–1	0.033***	0.029***	0.016***	0.047***
	(6.38)	(8.93)	(5.99)	(6.58)
Size: Micro	0.063***	−0.009*	0.005**	0.042***
	(25.65)	(−1.70)	(2.17)	(4.12)
Size: Small	0.094***	−0.010***	0.015***	0.049***
	(14.44)	(−3.07)	(3.62)	(13.38)
Size: Medium	0.170***	−0.028***	0.009***	0.056***
	(23.91)	(−4.49)	(4.17)	(8.96)
Business age: 6–10 years	0.003	−0.001	−0.005**	−0.024
	(0.32)	(−0.70)	(−2.14)	(−1.36)
Business age: 11–20 years	0.021*	−0.012***	−0.003	−0.046***
	(1.65)	(−5.45)	(−0.89)	(−6.30)
Business age: 20+ years	0.023***	−0.028***	−0.010***	−0.070***
	(2.73)	(−26.58)	(−2.59)	(−4.02)
Turnover change (stayed the same)t–1	−0.007**	−0.016***	−0.002	−0.017***
	(−2.02)	(−3.63)	(−0.70)	(−3.95)
Turnover change (increased)t–1	0.017**	−0.004	−0.001	−0.001
	(2.15)	(−1.13)	(−0.53)	(−0.41)
Profitt–1	0.007	−0.027***	−0.007**	−0.100***
	(1.13)	(−3.55)	(−2.47)	(−24.31)
Locationt: Urban area	−0.037***	−0.007***	−0.003**	−0.025***
	(−8.08)	(−3.73)	(−2.34)	(−8.40)
Family owned	0.041***	0.039***	0.002	0.012***
	(4.81)	(9.79)	(1.21)	(3.55)
Business plan	0.030***	0.001	0.007***	0.029***
	(29.59)	(0.50)	(3.81)	(5.74)
Fixed effects
Year / Regional / Industry FEs	YES	YES	YES	YES
Regional*Year / Industry*Year	YES	YES	YES	YES
N	9525	9525	9525	9525
Log pseudo-likelihood	−4293.720	−1853.036	−860.577	−3882.150
R^2	0.816	0.947	0.981	0.850
AIC	8593.440	3712.072	1727.155	7770.301
BIC	8614.925	3733.557	1748.640	7791.786
Non-event (failures = 0)	7774	9021	9343	8094
Event (successes = 1)	1751	504	182	1431

Notes: This table shows average marginal effects (AMEs) from a probit model of SMEs characteristics on the probability of using various sources of debt. All regressions include a constant term. The base categories for categorical variables are: zero employees (size), 0–5 years (business age), 18–30 years old (owner’s age), decreased (turnover change). All models include year, industry, regional, industry*year and regional*year fixed effects. Z-statistics adjusted for clustering at regional level are reported in parentheses. Statistical significance at the 10%, 5%, and 1% levels are showed by *, ** and ***.

4.1.2. Finance use by women-led and MEG-led social enterprises

By restricting the sample to social enterprises only, we can assess the differential importance of leadership diversity across social enterprises (women- and MEG-led) on the use of different forms of debt. The results presented in Table 6 suggest that relative to male-led counterparts, women-led social enterprises are less likely to rely on equity finance compared to male-led counterparts.¹⁶ However, MEG-leadership determines the use of certain types of funding. Specifically, relative to non-MEG counterparts, MEG-led social enterprises are 3.5% more likely to use bank overdrafts and 0.9% more likely to use equity finance. However, they are less likely to use: commercial mortgages (6.4%); factoring and invoice discounting (0.4%); government grants (12.95%); and leasing or hire purchase agreements (3.2%). Table 7 present findings in relation to the various forms of loans. The results suggest that relative to male-led counterparts, women-led social enterprises are 5.9% less likely to use loans from business partners, directors or owners. On the other hand, relative to non-MEG-led counterparts, MEG-led social enterprises are 3.8% more likely to use loans from mainstream financial intermediaries, and 3.6% more likely to use internal funding via loans from partners, directors or owners.

Table 6. Marginal effects of leadership diversity on use of finance by social enterprises.

	Bank overdraft	Commercial Mortgage	Credit Cards	Equity Finance	Factoring/Invoice discounting	Government or local authority grants	Leasing or hire purchase
Women-ledt–1	0.018	0.000	−0.010	−0.048***	0.001	−0.012	−0.019
	(1.25)	(0.01)	(−0.52)	(−4.36)	(0.14)	(−0.86)	(−1.36)
Minority ethnic-ledt–1	0.035***	−0.064***	−0.030	0.009*	−0.004*	−0.129***	−0.032**
	(2.77)	(−18.44)	(−1.47)	(1.81)	(−1.90)	(−10.35)	(−2.34)
Aims to growt–1	0.019	0.004	0.033**	0.008	0.039***	−0.012	0.070***
	(0.60)	(0.11)	(2.07)	(0.49)	(4.88)	(−0.45)	(5.80)
Size: Micro	−0.006	0.001	−0.005	−0.016	0.045**	0.060**	0.092***
	(−0.14)	(0.10)	(−0.14)	(−0.81)	(2.17)	(2.36)	(3.93)
Size: Small	−0.046***	0.042***	0.054	−0.006	0.037***	0.111***	0.204***
	(−3.13)	(14.85)	(1.04)	(−0.37)	(21.67)	(4.14)	(10.56)
Size: Medium	−0.008	0.183***	0.193***	−0.021	0.022	0.046**	0.284***
	(−0.77)	(8.18)	(4.10)	(−1.24)	(1.12)	(2.12)	(9.85)
Business age: 6–10 years	0.003	0.031***	0.019	−0.018	0.055***	0.072***	0.024
	(0.03)	(11.78)	(0.50)	(−0.91)	(3.42)	(5.35)	(1.16)
Business age: 11–20 years	0.085**	0.074***	0.166***	0.025	0.050***	0.110***	0.023
	(2.09)	(4.55)	(5.60)	(1.56)	(3.27)	(11.90)	(1.41)
Business age: 20+ years	0.115*	0.073***	0.163***	−0.032***	0.029*	0.124***	0.059***
	(1.93)	(10.39)	(7.50)	(−2.85)	(1.84)	(4.84)	(5.78)
Turnover change (stayed the same)t–1	−0.091***	−0.070***	−0.084	0.017	0.024***	−0.022	−0.098***
	(−5.77)	(−6.62)	(−1.55)	(0.99)	(4.33)	(−0.76)	(−3.11)
Turnover change (increased)t–1	−0.076***	−0.058***	−0.029	0.006	0.025**	0.015	−0.075***
	(−7.53)	(−7.57)	(−0.54)	(0.61)	(2.03)	(0.52)	(−4.65)
Profitt–1	−0.026	0.091***	−0.056	−0.003	0.036**	−0.085***	0.026
	(−0.85)	(7.08)	(−1.23)	(−0.49)	(2.08)	(−4.65)	(1.56)
Locationt: Urban area	−0.026*	0.001	0.019	0.017***	−0.006	−0.004	−0.037***
	(−1.92)	(0.12)	(0.98)	(4.57)	(−1.28)	(−0.22)	(−5.93)
Family owned	0.148***	0.002	0.036	−0.009	−0.003	−0.117***	0.005
	(13.05)	(0.26)	(1.14)	(−1.51)	(−0.33)	(−6.13)	(0.18)
Business plan	0.132***	−0.012	0.084***	0.041**	0.045**	0.043	0.065***
	(5.96)	(−0.81)	(17.75)	(2.50)	(2.21)	(1.50)	(4.05)
Fixed effects
Year / Regional / Industry FEs	YES	YES	YES	YES	YES	YES	YES
Regional*Year / Industry*Year	YES	YES	YES	YES	YES	YES	YES
N	781	781	781	566	773	773	781
Log pseudo-likelihood	−437.527	−209.576	−487.979	−59.980	−170.434	−279.709	−363.647
R^2	0.714	0.903	0.661	0.972	0.935	0.850	0.787
AIC	881.053	425.151	981.957	125.961	346.868	565.417	733.293
BIC	895.035	439.133	995.939	138.976	360.818	579.368	747.275
Non-event (failures = 0)	558	704	480	550	723	643	606
Event (successes = 1)	223	77	301	16	50	130	175

Notes: This table shows average marginal effects (AMEs) from a probit model of social enterprises’ characteristics on the probability of using various sources of debt. The sample is restricted to SMEs which are classified as social enterprises. All regressions include a constant term. The base categories for categorical variables are: zero employees (size), 0–5 years (business age), 18–30 years old (owner’s age), decreased (turnover change). All models include year, industry, regional, industry*year and regional*year fixed effects. Z-statistics adjusted for clustering at regional level are reported in parentheses. Statistical significance at the 10%, 5%, and 1% levels are showed by *, ** and ***.

Table 7. Marginal effects of leadership diversity on use of finance by social enterprises.

	Loan from a bank, building society or other financial institution	Loan from family/friend	Loan from business partner/director/owner
Women-ledt–1	0.011	0.017	−0.059***
	(0.91)	(1.62)	(−4.34)
Minority ethnic-ledt–1	0.038**	0.012	0.036***
	(2.34)	(0.72)	(3.12)
Aims to growt–1	0.021	0.054*	0.046
	(0.82)	(1.84)	(1.14)
Size: Micro	0.038*	−0.045***	−0.061**
	(1.83)	(−4.42)	(−2.05)
Size: Small	0.108***	−0.028***	−0.067**
	(4.21)	(−3.70)	(−2.10)
Size: Medium	0.184***	n.e	−0.073
	(4.82)	(.)	(−1.57)
Business age: 6–10 years	−0.090*	0.008	−0.026
	(−1.80)	(0.24)	(−0.52)
Business age: 11–20 years	0.022	0.031	−0.012
	(0.34)	(0.72)	(−0.43)
Business age: 20+ years	−0.022	0.008	−0.055
	(−0.32)	(0.37)	(−1.16)
Turnover change (stayed the same)t–1	−0.078***	−0.040***	−0.047***
	(−2.68)	(−2.99)	(−4.54)
Turnover change (increased)t–1	−0.032	−0.027**	−0.035***
	(−0.84)	(−2.28)	(−5.54)
Profitt–1	0.051	−0.014	−0.038**
	(0.90)	(−1.07)	(−2.25)
Locationt: Urban area	−0.041*	−0.022*	−0.045
	(−1.70)	(−1.86)	(−1.40)
Family owned	0.044*	0.051***	0.007
	(1.73)	(7.59)	(0.43)
Business plan	0.040**	0.007	0.061***
	(2.35)	(0.90)	(3.72)
Fixed effects
Year / Regional / Industry FEs	YES	YES	YES
Regional*Year / Industry*Year	YES	YES	YES
N	781	624	781
Log pseudo-likelihood	−308.939	−124.554	−237.868
R^2	0.846	0.941	0.898
AIC	623.878	253.108	481.736
BIC	637.860	261.980	495.717
Non-event (failures = 0)	655	587	700
Event (successes = 1)	126	37	81

Notes: This table shows average marginal effects (AMEs) from a probit model of social enterprises’ characteristics on the probability of using various sources of debt. The sample is restricted to SMEs which are classified as social enterprises. Loans from a peer-to-peer platform has been excluded because of lack of data. All regressions include a constant term. The base categories for categorical variables are: zero employees (size), 0–5 years (business age), 18–30 years old (owner’s age), decreased (turnover change). All models include year, industry, regional, industry*year and regional*year fixed effects. The term ‘n.e’ stands for ‘not estimable.’ This condition arises when the outcome variable can be perfectly predicted from a subset of the data, or when there are insufficient observations to compute the marginal effects. Z-statistics adjusted for clustering at regional level are reported in parentheses. Statistical significance at the 10%, 5%, and 1% levels are showed by *, ** and ***.

4.2. Access to finance by social enterprises

In this section, we present the results from a Heckman probit model with sample selection (Van de Ven and Van Praag 1981). In this empirical setting, the selection equation in Table 8 relates to the probability of applying for finance (demand) and the outcome equation relates to the probability of obtaining finance conditional upon having applied for finance.

Table 8. Access to finance by social enterprises.

	Model 1: Bank overdrafts		Model 2: Commercial mortgage		Model 3: Credit Cards		Model 4: Government or local authority grants or schemes		Model 5: Loans from banks building societies, etc.
	Selection	Outcome	Selection	Outcome	Selection	Outcome	Selection	Outcome	Selection	Outcome
Social enterprise	−0.109***	0.026	−0.017	0.183***	0.008	0.091***	0.093***	0.074***	−0.048	0.055***
	(-4.93)	(0.61)	(-1.04)	(10.31)	(0.45)	(4.49)	(3.00)	(2.69)	(-1.26)	(3.49)
Women-led t-1	−0.034	−0.055*	0.009	0.009	0.039	−0.117***	−0.003	0.009	0.054**	−0.045***
	(−1.35)	(−1.86)	(0.88)	(0.45)	(1.12)	(−7.34)	(−0.33)	(0.30)	(2.40)	(−3.09)
Minority ethnic-led t-1	−0.026	−0.167***	−0.006	0.940*	−0.080***	0.529***	−0.060***	1.500***	0.018	−0.069
	(−1.59)	(−4.94)	(−0.22)	(1.65)	(−2.61)	(5.76)	(−5.56)	(7.46)	(0.39)	(−1.44)
Aims to grow t-1	−0.021	−0.064***	0.007	0.034	−0.019**	0.002	0.022***	−0.014	0.040	0.013
	(−1.49)	(−4.69)	(0.77)	(0.65)	(−2.20)	(0.09)	(4.66)	(−0.26)	(1.45)	(0.34)
Size: Micro	0.030**	−0.035	0.014	−0.126**	0.006	−0.021*	0.039***	0.456***	0.071***	0.065
	(2.02)	(−1.63)	(1.57)	(−2.29)	(0.18)	(−1.82)	(4.76)	(7.36)	(4.02)	(1.58)
Size: Small	−0.008	0.010	0.033***	−0.092***	0.012	0.062***	0.000	0.491***	0.038***	0.120*
	(−0.51)	(0.33)	(3.48)	(−2.65)	(0.27)	(3.14)	(0.03)	(15.34)	(3.46)	(1.84)
Size: Medium	−0.027*	0.046	0.072***	0.060***	−0.005	0.050	0.013	0.500***	0.118***	0.174***
	(−1.92)	(1.15)	(20.65)	(3.99)	(−0.15)	(1.61)	(1.53)	(4.08)	(5.65)	(5.12)
Business age: 6 – 10 years	0.014	−0.007	−0.037**	0.201***	0.019	0.012	0.066***	−0.166***	0.012	−0.100***
	(0.56)	(−0.13)	(−2.05)	(41.38)	(0.81)	(0.13)	(4.59)	(−3.10)	(0.28)	(−4.34)
Business age: 11 – 20 years	0.040	0.065	−0.028*	0.067***	0.060***	0.100***	0.046***	−0.312***	−0.011	−0.094**
	(1.35)	(1.33)	(−1.75)	(2.64)	(4.04)	(2.62)	(2.93)	(−2.90)	(−0.51)	(−2.28)
Business age: 20+ years	0.100***	0.086**	0.007	0.108***	0.035	0.063*	0.027**	−0.163*	0.015	−0.043**
	(13.59)	(1.98)	(0.31)	(7.50)	(1.47)	(1.82)	(2.38)	(−1.81)	(0.42)	(−2.47)
Turnover change (stayed the same) t-1	0.007	−0.019	−0.004	−0.001	0.012	0.001	0.003	0.059	−0.031	0.128***
	(0.67)	(−0.95)	(−0.28)	(−0.04)	(1.32)	(0.03)	(0.18)	(0.47)	(−0.91)	(11.88)
Turnover change (increased) t-1	0.046***	−0.032	−0.009	−0.033	0.016	−0.020	0.005	−0.121*	−0.004	0.037**
	(4.60)	(−0.66)	(−1.55)	(−1.27)	(1.07)	(−1.36)	(0.42)	(−1.83)	(−0.15)	(2.33)
Profit t-1	−0.040	0.082**	0.012	0.285***	−0.028*	−0.025***	−0.033***	0.130	−0.018	0.142***
	(−1.41)	(2.44)	(0.77)	(9.69)	(−1.71)	(−7.00)	(−6.87)	(1.01)	(−0.52)	(7.42)
Location t: Urban area	−0.051*	−0.036	0.009	0.027	0.008	0.007	−0.020	0.015	−0.016*	−0.040***
	(−1.92)	(−1.09)	(0.79)	(1.37)	(0.51)	(0.22)	(−1.39)	(0.64)	(−1.86)	(−2.61)
Family owned	0.002	0.027*	0.041***	−0.070	0.008	−0.094***	−0.043**	−0.087	0.049	−0.005
	(0.06)	(1.88)	(5.40)	(−0.72)	(0.27)	(−3.82)	(−2.35)	(−1.02)	(1.44)	(−0.30)
Business plan	−0.032	0.025	0.011	−0.038***	−0.035***	−0.000	0.047***	0.072	−0.027**	0.011
	(−0.83)	(1.07)	(1.39)	(−2.97)	(−3.78)	(−0.00)	(8.63)	(1.32)	(−2.46)	(0.67)
Business advice	0.041***	−0.006	0.020***	0.016	0.020**	0.004	0.015*	0.013	0.025*	0.004
	(4.47)	(−0.17)	(4.82)	(0.67)	(2.04)	(0.19)	(1.72)	(0.59)	(1.69)	(0.68)
Athrho	0.366		−1.079		−0.387		−0.485		−0.342
	(0.18)		(−0.73)		(−0.21)		(−0.54)		(−0.83)
Ρ	0.351		−0.793		−0.369		−0.450		−0.329
N	1283		1281		1293		1288		1263
Selected	454		91		169		88		493
Non-selected	829		1190		1124		1200		770
Log pseudo-likelihood	−967.338		−328.007		−514.103		−306.379		−1045.141
Wald test of indep. Eqns (ρ = 0)	0.032		0.533		0.042		0.289		0.687
Prob > chi2	0.859		0.466		0.837		0.591		0.407

Notes: This table present the marginal effects from a Heckman probit model with sample selection. The selection equation relates to the probability of applying for finance (demand). The outcome equation relates to the probability of obtaining finance conditional on having applied for finance. All regressions include a constant term. The exclusion restriction used in the selection equation is whether the firm used business advice in the last 12 months. The base categories for categorical variables are: zero employees (size), 0-5 years (business age), 18–30 years old (owner’s age), decreased (turnover change). All models include industry and regional fixed effects, except for the outcome equation in Model 4 where regional effects were excluded to achieve convergence. When $\rho =0$, there is no evidence of selection bias; and thus, the outcome and selection equations are independent, making the estimation of the selection model unnecessary. However, since the model is estimated by maximum likelihood (ML), ρ is not directly estimated. Instead, the Heckprobit routine directly estimates a nonlinear transformation of ρ (athrho) defined as: $\text{athrho}=\frac{1}{2}\text{ln}\left(\frac{1+\rho }{1-\rho }\right)$. A significant athrho indicates the presence of selection bias in the model. Models 1, 3, and 4 include region, sector, year and sector*year fixed effects, while models 2 and 5 include sector, region and year fixed effects. Z-statistics adjusted for clustering at the regional level are reported in parentheses. ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively.

The results presented in Table 8 complement the findings provided in the previous section regarding the use of specific forms of debt. The results suggest that compared to mainstream SMEs, social enterprises are 10.9% less likely to apply for bank overdrafts, but 9.3% more likely to apply to government or local authority grants or schemes. Interestingly, conditional upon application, social enterprises have an 18.3% greater chance of success obtaining a commercial mortgage, 9.1% greater chance in securing credit card funding, 7.4% higher probability of securing a government grant, and 5.5% higher probability of securing loans from banks compared to mainstream SMEs.

Finally, we assess whether leadership diversity within social enterprises has an influence on access to finance. Here, we use results reported in Table 8 to compute average marginal effects for women-led and MEG-led firms conditional upon being social enterprises, while adjusting for all other covariates. The results reported in Figure 4 focus on applications, and suggest that women-led social enterprises are 5.2% more likely to apply for loans from a mainstream financial intermediary. However, MEG-led social enterprises are 8.3% less likely to apply for credit card funding and 11% less likely to apply to government or local authority grants.

Figure 4. Conditional marginal effects of leadership diversity of social enterprises on demand for main finance sources.

Notes: This Figure shows average marginal effects (AMEs) from results reported in Table 8 (selection equation) for women-led (Figure 4A) and MEG-led businesses (Figure 4B) conditional on being social enterprises, while adjusting for all other covariates. This figure uses a horizontal layout in which sources of funding (Models 1–5 in Table 8) are placed on the Y-axis and the estimated AMEs and their (99%, 95% and 90%) confidence intervals are plotted along the X-axis.

In terms of the outcome of funding applications, the results in Figure 5 suggest that women-led social enterprises are 3.7% less likely to secure loans from mainstream financial intermediaries compared to male-led counterparts. The results for MEG-led social enterprises suggest that this group of businesses are 15% less likely to secure funding via bank overdrafts. However, conditional on application, MEG-led social enterprises present the highest probability of securing funding from government or local authority grants compared to non-MEG-led counterparts.

Figure 5. Conditional marginal effects of leadership diversity of social enterprises on access to main finance sources.

Notes: This Figure shows average marginal effects (AMEs) from results reported in Table 8 (selection equation) for women-led (Figure 5A) and MEG-led businesses (Figure 5B) conditional on being social enterprises, while adjusting for all other covariates. This figure uses a horizontal layout in which sources of funding (Models 1–5 in Table 8) are placed on the Y-axis and the estimated AMEs and their (99%, 95% and 90%) confidence intervals are plotted along the X-axis. The results for credit card have been excluded because it was not possible to estimate their AMEs.

5. Discussion and theoretical implications

This paper contributes to the literature on access to finance and financial resource mobilisation of social enterprises. It also provides new insights regarding the impacts of leadership diversity on access to finance. Our results show distinct patterns in funding strategies of social enterprises compared to mainstream SMEs. This aligns with prior evidence which suggests that the financing sources of social enterprises differ from other businesses (Austin, Stevenson, and Wei-Skillern 2006; Dees 1998; Lumpkin et al. 2013). Our results suggest that social enterprises have a lower propensity to use traditional banking products (such as overdrafts and loans), and instead rely on factoring/invoice discounting and grants funding provided by central government and local authorities. This suggests that mainstream lenders do not offer financial products appropriate to the needs of social enterprises (Spiess-Knafl and Scheck 2019).

Our results also shed light on the challenges faced by women-led and MEG-led social enterprises. We find that women-led social enterprises are less likely to use equity finance and loans from business partners, indicating potential barriers in accessing a broad range of funding options. MEG-led social enterprises exhibit a varied pattern in the use of funding sources, with less reliance on the use of commercial mortgages, factoring and leasing and government grants, but more use of bank overdrafts, equity finance and loans from financial intermediaries and business partners/directors and owners. Our results suggest the need for more tailored support from government in terms of grant funding for MEG-led social enterprises. Overall, the observed disparities in the use of finance for women-led and MEG-led social enterprises calls for policy interventions to ensure that social enterprises have adequate access to appropriate funding regardless of leadership.

An important contribution of this study is the empirical analysis of both the demand-side (funding applications) and supply-side (funding outcomes) aspects related to access to finance. Our empirical approach allows us to assess where finance access gaps occur. Relative to commercial SME counterparts, our results suggest that social enterprises are less likely to apply for bank overdrafts, but more likely to apply for government grants. However, conditional on applying for different forms of finance, social enterprises have a higher success rate in obtaining funding via commercial mortgages, credit cards, bank loans, and government grants. Further analyses of women-led and MEG-led social enterprises suggest that leadership diversity matters in terms of both the demand and supply of funding. Women-led social enterprises apply more for bank loans, but have a lower success rate compared to male-led enterprises. MEG-led social enterprises apply less for credit card funding and government grants. However, upon application, they are less likely to receive bank overdrafts, but more likely to receive grants compared to non-MEG-led counterparts.

The findings of our study, when viewed through the lens of banking and finance theories, can offer insights that can be extrapolated to other cases involving social enterprises and their access to finance, but also provide a framework for predicting and understanding the problems faced by social enterprises in the context of access to finance. Our finding that social enterprises rely less on traditional banking products and more on alternative financing methods such as factoring and government grants can be understood through credit rationing theories. Stiglitz and Weiss (1981) argue that lenders lack perfect information regarding borrowers’ risk, leading to higher interest rates and rationing of credit for SMEs, who face higher information opacity compared to larger firms. Berger and Udell (1998) argue that information opacity is more severe in SMEs than in large firms, affecting capital structure decisions and debt financing. Cowling (2010) emphasise the role of information-based problems that limit the supply of bank credit to SMEs. These theories suggest that traditional financial markets could fail to serve certain businesses like social enterprises adequately due to perceived higher risks, information opacity or lower profitability.

The challenges faced by women-led social and MEG-led businesses in accessing a broad range of funding options reflect the outcomes of asymmetric information. Lack of funding can therefore occur due to poor information flows associated with moral hazard and adverse selection issues (Diamond 1984). Asymmetric information also leads to credit rationing, with smaller and younger firms finding it more difficult to access finance due to increased bank screening costs (Finnegan and Kapoor 2023; Masiak et al. 2019). This insight can be extrapolated to social enterprises where information asymmetry can hinder the financial accessibility for these types of non-traditional business models. In summary, the application of these theories to our findings provides a framework for understanding and predicting the financial behaviours and challenges of social enterprises.

6. Conclusions

Social enterprises are a unique form of organisation pursing economic, ethical, social and environmental goals. As such, their respective commercial activities intersect with the significant social, ethical and environmental challenges facing society today. In this study, we use 2016–2019 waves of the LSBS survey to investigate access to finance issues faced by UK social enterprises. The findings of an extensive econometric investigation suggest that social enterprises are less likely to apply for bank overdrafts, but are more likely to apply to government grants compared to mainstream SMEs. However, upon applying for funding, social enterprises are more likely to receive commercial mortgages, credit card financing, government grants and loans from mainstream financial intermediaries compared to commercial SME counterparts.

In terms of leadership diversity of social enterprises, our results suggest that women-led social enterprises are less likely to use equity finance and loans from business partners/directors/owners. We also find that MEG-led social enterprises rely less on commercial mortgages, factoring/invoice discounting, government grants and leasing or hire purchase forms of finance compared to non-MEG-led counterparts, but are more likely to use bank overdraft, equity finance, loans from mainstream financial intermediaries (e.g. a bank, building society or other financial institution) or loans from a business partner/director/owner. Considering the effect of leadership diversity of social enterprises on their access to finance, our results suggest that women-led social enterprises are more likely to apply for loans from a bank but, conditional on application, less likely to receive funding compared to male-led social enterprises. Our results also show that MEG-led social enterprises are less likely to apply for credit cards and government grants; and conditional on application, less likely to get a bank overdraft facility. However, conditional on application, they MEG-led social enterprises exhibit the largest probability of securing funding from a government or local authority grants compared to their non-MEG-led counterparts.

The insights presented in this paper have broad relevance beyond the UK. Social enterprises in other countries are likely to face similar financing constraints, given that their hybrid business models do not fit neatly into traditional funding frameworks designed for commercial businesses or non-profits. These constraints may be even greater in countries with less developed social enterprise ecosystems and supportive government policies. The prevalence of women-led and minority ethnic-led social enterprises in the UK highlights the importance of diversity and inclusion for this sector. Other countries may similarly benefit from enabling greater access to finance for social enterprises led by women or ethnic minorities. The results of this study suggest that there is a need for tailored financial instruments and support programmes aligned with the business models of social enterprises. The findings provide relevant lessons applicable across sectors and regions, which can be also transferable to similar organisations facing specific financing constraints. These findings underscore the necessity of a holistic approach to alleviate access barriers for social enterprises, including financing mechanisms, educational initiatives, and policy interventions. In summary, improving access to funding for social enterprises requires a recognition of their unique hybrid nature and leadership diversity.

In terms of future research, there is a need for more quantitative and qualitative research in order to gain deeper insights into the specific funding needs and challenges faced by social enterprises. Building on the foundational insights provided by our current research, future studies could focus on borrower discouragement among social enterprises, particularly those led by women and MEGs. Complementing recent work related to borrower discouraged SMEs in the UK (Brown, Liñares-Zegarra, and Wilson 2022), future research could employ specific survey questions from the LSBS to delve deeper into the dynamics of borrower discouragement for social enterprises. Such investigations could explore the specific causes and effects of borrower discouragement, especially in the context of accessing finance and across underrepresented leadership groups within this sector. Additionally, impact evaluations of specific funding sources tailored to social enterprises at various stages of development could further elucidate the effectiveness of current policies. Future research should explore in more detail the causal mechanisms behind these differences and explore effective strategies to ensure social enterprises are able to obtain the required finance to fulfil their organisational objectives. This is important for the sustained growth and impact of these organisations in addressing social and environmental challenges.

Overall, the results presented in this study have important implications for public policy by providing valuable information for organisations and other key stakeholders introducing or monitoring interventions, or offering financial support to UK social enterprises. Social enterprises face specific barriers to access to finance, which differ from those encountered by the wider SME sector. Having a business model where profits are used to achieve social, environmental and ethical goals appears to exacerbate many of the access to finance barriers facing SMEs. This issue particularly affects MEG and women-led social enterprises. More public policy support is needed to bridge the existing knowledge and funding gaps. By providing tailored financial assistance that aligns with their specific mission, business model, industry, and stage of development, these enterprises can access the resources they need. This will enable them to fulfil their unique potential.

Acknowledgements

We are grateful to Barbara Casu, Laura Chiaramonte, Douglas Cumming, Franco Fiordelisi, Claudia Girardone, Sotirios Kokas, Alessandra Mongiardino and other participants at the Workshop on Social Responsibility, Banks and Markets During the COVID-19 Crisis and Beyond for useful comments and suggestions. We gratefully acknowledge the ESRC-funded Enterprise Research Centre and the Department for Business, Energy and Industrial Strategy (BEIS) for funding for this study. In order to meet institutional and research funder open access requirements, any accepted manuscript arising shall be open access under a Creative Commons Attribution (CC BY) reuse licence with zero embargo.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Enterprise Research Centre at Warwick.

Notes on contributors

José M. Liñares-Zegarra

Dr. José M. Liñares-Zegarra is a Senior Lecturer (Associate Professor) in Finance at Essex Business School (UK) and an affiliated researcher at the Centre for Responsible Banking & Finance (CRBF). He also holds an honorary position at the University of St Andrews, Scotland. With more than 15 years of experience in teaching and research, José's specializations include banking, financial technology (FinTech), and entrepreneurship. His research has been published in top-tier international journals such as the British Journal of Management, Journal of Corporate Finance, European Journal of Finance, and Small Business Economics, among others. Prior to his current roles, José held positions at various national and international institutions, including the Federal Reserve Bank of Boston, the University of Alicante, the University of Illinois, Florida State University, and the Institute of European Finance at Bangor University.

John O. S. Wilson

Professor John O. S. Wilson is Professor of Banking & Finance and Head of the Department of Finance at the Business School, University of St Andrews. He is also the Director of the Centre for Responsible Banking & Finance. In the period June 2011 to April 2012, John served as a full member of a Commission on Credit Unions established by the Irish Government. In 2018, John delivered evidence on the impact of Brexit on UK small and medium-sized enterprises to the House of Lords EU Internal Markets Committee. In 2020, John was appointed as a full panel member to the 2021 Research Evaluation Framework (REF) Sub-panel 17: Business and Management Studies.

Notes

1 BEIS (2020), drawing on data from the LSBS, shows that the primary barriers SMEs face in seeking external finance have remained consistent from 2015 to 2019. These include uncertainty about economic conditions, the fear of rejection, and SMEs' reluctance to take on additional risks, which continue to be the top deterrents.

2 Based on the LSBS classification (see Figure 2), social enterprises are classified as enterprises that have: identifiable social/environmental goals; generate income from trading activities (i.e. engage in entrepreneurial activity); and use surplus/profit to further social/environmental goals. We refer the reader to Section 3.2 for a detailed explanation of the methodology used by the LSBS to identify social enterprises.

3 Estimates of total employment accounted for by UK social enterprises range from 1.8 million (DCMS/BEIS 2021) to 2.3 million (Social Enterprise UK 2023) people. The figures reported by Social Enterprise UK include cooperatives. Therefore, the aggregate employment figures for social enterprises excluding cooperatives are approximately 1.85 million and 2.3 million including cooperatives.

4 Mainstream SMEs, as referred to in this paper, are businesses within the small business population that are neither traditional non-profits (mostly charities, which are excluded from the sample) nor social enterprises (see Footnote 9 and Section 3.3 for details).

5 "For the most part" acknowledges the existence of large social enterprises as part of the UK business population. Cabinet Office (2013) reports 6455 large UK enterprises in 2012. If their proportion of social enterprises mirrored SMEs, there could be an estimated 400–900 additional large social enterprises. This indicates a possible, though not thoroughly surveyed, presence of sizable social enterprises.

6 Typologies and definitions of social enterprises are numerous and varied. Extensive early discussions and taxonomies of social enterprises can be found in Austin, Stevenson, and Wei-Skillern (2006). Other useful discussions regarding definitions and typologies of social enterprises include Zahra et al. (2009), Dacin, Dacin, and Matear (2010), Teasdale (2012), Doherty, Haugh, and Lyon (2014). Saebi, Foss, and Linder (2019) provide a more general overview of the salient literature.

7 The levelling up agenda is a government initiative designed to tackle persistent economic and social inequalities prevalent across the UK. This is underpinned by four overarching objectives aimed at enhancing productivity, wages, employment and living standards via growth in private sector economic activity; improving access to and the provision of public services; restoring pride to community; and providing funding and support for local empowerment.

8 Brown, Liñares-Zegarra, and Wilson (2019) suggest that firms located in peripheral geographical areas have greater usage of credit cards relative to counterparts located in ‘core’ location.

9 Traditional non-profits, mainly charities, are excluded from the sample due to their distinct income sources, legal structure, and governance compared to other SMEs under study. They rely heavily on donations, grants, and non-trading income sources, generating less than 50% of their income from trading activities. This differs from mainstream SMEs and social enterprises, which primarily generate income from trading activities. Including traditional non-profits could skew the analysis due to their fundamentally different financial structures and revenue models. Additionally, their legal status and governance structures can differ significantly from those of mainstream SMEs or social enterprises, affecting their operations, accountability mechanisms, and decision-making processes. Excluding traditional non-profits ensures a more homogeneous and comparable sample for analysis, focusing on entities with similar operational and financial characteristics.

10 All LSBS Questionnaires for the 2016–2019 waves and methodological explanations are available online on the Department for Business and Trade website at https://www.gov.uk/government/collections/small-business-survey-reports. Table 1 provides definitions of the variables used and specific mnemonics, which can be used to identify specific survey questions used in the present study.

11 Typologies and definitions of social enterprises are numerous and varied. Extensive early discussions and taxonomies of social enterprises can be found in Austin, Stevenson, and Wei-Skillern (2006). Other useful discussions regarding definitions and typologies of social enterprises include Zahra et al. (2009), Dacin, Dacin, and Matear (2010), Teasdale (2012), Doherty, Haugh, and Lyon (2014). Saebi, Foss, and Linder (2019) provide a more general overview of the salient literature.

12 The lower number of observations for the dummy variable identifying social enterprises and mainstream SMEs in Table 2 is because the LSBS's ‘Social Enterprises’ module was only included in the 2017 and 2019 Survey Waves. However, SME characteristics were recorded in every survey wave from 2016 to 2019, leading to this discrepancy in the number of observations.

13 The use of firm-level fixed effects in our binary regression models posed significant challenges due to the models’ non-linear nature. This caused issues with perfect prediction and severe convergence problems during estimation, including perfect collinearity with other time-invariant predictors. These challenges highlight the need for further research in this area as more comprehensive longitudinal data becomes available.

14 Marginal effect estimates capture how the probability of the dependent variable changes as the predictor changes. The marginal effect for a continuous independent variable is the partial derivative of the event probability with respect to the variable of interest. For a binary independent variable, this is the change in probability when the variable of interest changes from 0 and 1.

15 Table 2 provides a comprehensive view of the raw dataset from the LSBS database for transparency. The smaller sample sizes used in Tables 4–8 result from the availability of data for key variables. This includes the use of a social enterprises dummy variable and lagged variables aimed at addressing endogeneity, which necessitate data from consecutive survey periods, thus impacting size of the final sample. The Appendix Tables A1, A2, and A3 provide detailed descriptive statistics of this adjusted sample for Tables 4–8.

16 Given the less frequent use of equity finance among social enterprises, the results in Table 6 include only 16 ‘successful’ observations (i.e. where the dependent variables takes the value equals to 1). Consequently, we are cautious in drawing strong inferences from these particular findings.