Microfinance and Small Business Development in a Transitional Economy: Insights from Borrowers’ Relations with Microfinance Organisations in Kazakhstan

Abstract

Microfinance is seen as an important vehicle for developing small businesses in developing and transitional economies despite the relative absence of supporting research. We use mixed methods to offer a nuanced empirical exploration of the relationship between microfinance and everyday entrepreneurial practice(s) in Kazakhstan. As in many transitional contexts, ‘unbankable’ borrowers here operate in a vibrant informal sector, face high degrees of uncertainty, and retain a strong distrust of a corrupt/predatory state. Our data-based methodology for analysing borrowers’ diverse relationships with microfinance organisations (MFOs) generates insights into their multiple pathways to business development. Both ‘outreac\h’ and ‘commercialised’ MFOs sustain micro-flows of resources that are critical for everyday entrepreneurs who need to finance ongoing consumption and contingencies whilst also (and by) building up their small businesses. Microfinance use did not promote formalisation or impersonalised banking relationships. Instead, MFOs focused primarily on repayment, clients’ businesses remained partially formalised or unregistered across all stages of growth and the lending relationships preferred by Private MFOs and borrowers were highly personalised. Consequently, we call for assumptions about how microfinance can (and should) drive small business development need to be rethought for transitional contexts.

KEYWORDS:

• Microfinance
• entrepreneurial practice
• transitional economy
• Kazakhstan

1. Introduction

Microfinance is widely viewed as a substitute for informal sources of finance and as fostering entrepreneurial growth and poverty alleviation in developing and transitional economies (Bruton, Ahlstrom, & Si, 2015; Parker, 2009; Weber & Ahmad, 2014; Zhao & Lounsbury, 2016). Despite its enthusiastic promotion and widespread adoption, its impact on entrepreneurial growth and poverty alleviation is contested (Banerjee, Duflo, Glennerster, & Kinnan, 2015; Bateman & Chang, 2012), and increasingly microfinance is criticised for encouraging subsistence, necessity entrepreneurship, being unable to generate sustainable entrepreneurship, charging high interest rates and exacerbating vulnerabilities and indebtedness (Banerjee & Jackson, 2017; Bruton et al., 2015; Chen, Chang, & Bruton, 2017; Smith, Judge, Pezeshkan, & Nair, 2016; Webb, Bruton, Tihanyi, & Ireland, 2013). A key reason for this inconclusiveness is the under-development of research on entrepreneurial practice(s) in relation to microfinance in specific developing or transitional contexts.

There has been very limited exploration of microfinance borrowers, their motivations and most importantly, their use of loans over time (Canales & Greenberg, 2016; Chliova, Brinckmann, & Rosenbusch, 2015; Shahriar & Garg, 2017), but emerging scholarship draws attention to varying relationships between entrepreneurial growth, microfinance and institutional formalisation (Si, Ahlstrom, Wei, & Cullen, 2020; Sutter, Bruton, & Chen, 2019). Calls for more nuanced attention to borrower diversity and agency in situ (Xheneti, Madden, & Thapa Karki, 2019) are growing and query the teleological idea of institutional formalisation in which (all) economies (and entrepreneurs) seek to move from mostly informal to mostly formalised economic activities and that this is good and desirable (Williams & Round, 2007). Scholars increasingly believe that a hyper-concern with formal firms and for formalisation is misplaced, missing both the heterogeneity of ‘everyday entrepreneurship’ and the ‘broader context of reasons, purposes and values for why and how entrepreneurship emerges’ (Welter, Baker, Audretsch, & Gartner, 2017, p. 311).

We build on these starting points to explore borrowers’ relationships with microfinance organisations (MFOs) in support of their dynamic entrepreneurial activities in the transitional economy of Kazakhstan. Understanding this is important for transitional economies where small business development is integral to successfully building a market economy (Bliss & Garratt, 2001; Davis, 2016; Low, 2006). As with other similar economies, the majority of MFO borrowers in Kazakhstan are entrepreneurs in the informal economy who are overlooked by commercial banks, and hitherto have relied on informal moneylenders (Subalova, Al-Dajani, & Bika, 2015). In these contexts, where a large proportion of the workforce are by necessity self-employed, the informal sector plays a vital role in poverty alleviation, economic growth and in fostering the new market economy. As a result, microfinance has become an important policy tool for transitional economies in stimulating small business sector development, albeit without a clear empirical understanding of how it drives entrepreneurial behaviour.

We present a detailed mixed-methods case study of Almaty and Almatinskaya Districts of Kazakhstan where microfinance is thriving. Our objective is to investigate the relationship between borrowers’ ‘everyday’ entrepreneurship (Welter et al., 2017) and MFOs in this transitional economy and to consider what this means for the impact of microfinance on entrepreneurial growth. We begin by reviewing emerging thinking about research on the relationship between microfinance and entrepreneurial practice(s), arguing for the importance of a methodology that asks questions about everyday entrepreneurs and their various pathways in ways that are strongly embedded in context (Jackson, Helfen, Kaplan, Kirsch, & Lohmeyer, 2019, p. 25).

2. Re-orientating research on microfinance and entrepreneurial practice

Despite an increasing volume of research, expertly reviewed by Chen et al. (2017), the evidence on the entrepreneurial outcomes of microfinance is inconsistent. The reasons for this lie partly in a series of persistent dualities that have shaped microfinance and entrepreneurial thinking and research, and partly in a failure to understand the importance of the institutional context for making sense of what the expansion of microfinance and related entrepreneurial activity means over time for entrepreneurial growth and poverty alleviation in specific places. We briefly review both weaknesses below and summarise the calls from emerging research to develop a more inquisitive and nuanced approach towards understanding microfinance and entrepreneurial practice(s) in different contexts.

Most research on microfinance focuses on either repayment rates and the sustainability of MFOs or the effectiveness of group lending practices and borrower outcomes (Khavul, 2010, p. 63). Both offer overly dualistic perceptions of the differences amongst MFOs, as commercial or outreach (Hoque, Chishty, & Halloway, 2011; Mcintosh & Wydick, 2005; Morduch, 2000), and borrowers, as poor and ‘necessity driven’ (Karlan & Zinman, 2012; Newman, Schwarz, & Ahlstrom, 2017; Newman, Schwarz, & Borgia, 2014) or as ‘would-be’ entrepreneurs who are ‘opportunity driven’ (Chliova et al., 2015; Shahriar & Garg, 2017; Siwale & Ritchie, 2012). In practice, many MFOs serve both kinds of customers, and the boundaries between them are fuzzy (Khavul, Chavel, & Bruton, 2013) whilst many borrowers’ motivations move between, and may straddle, different categories over time as circumstances require or allow (Beck, Aquilera, & Schintz, 2018; Williams, 2008). Current research heeds longstanding calls to move away from asking which borrowers are ‘real’ entrepreneurs (Gartner, 1988; Ramoglou, Gartner, & Tsang, 2020) to instead investigating how borrowers use microfinance in their ‘everyday entrepreneurship’ (Welter et al., 2017).

Borrowers have often been approached as the rule-takers in the lending relationship (Schäfer, Siliverstovs, & Terberger, 2010), but it is increasingly clear that MFOs have to take borrower preferences into account in order to thrive and grow (Baraton & Léon, 2019) and that borrowers strategically negotiate between different MFOs and different lending arrangements in order to meet their own (changing) interests (Cohen, 2002). Extensive research on lending to women in developing countries shows how men and women circumvent the empowerment aspirations of donors and MFOs to meet their gendered interests (Cervantes, Montoya, & Ponce, 2017). In sum, there is a need for researchers to give ‘expanded attention to agency’ (Sutter et al., 2019, p. 209). Accordingly, we see microfinance clients operating largely (but not necessarily wholly) within the informal economy as agentic ‘entrepreneurial borrowers’ whose motivations for borrowing are dynamic, contextually embedded and may often combine daily needs, working capital and business investments.

This perspective involves raising questions about orthodox assumptions about how microfinance ‘successfully’ enables ‘would-be’ entrepreneurs to develop their businesses. Microfinance is widely understood to drive small business development by promoting enterprise growth, registration, and formalisation, including socialisation into standardised lending relations (Akula, 2008; Chakrabarty & Bass, 2014; Ranabahu & Barrett, 2020; Xheneti et al., 2019). However, presuming that formalisation is the desired end for micro-entrepreneurs cannot explain spatio-temporal variations in outcomes (Williams, Martinez–Perez, & Kedir, 2017), neglects the importance of the informal economy as a source of entrepreneurialism and livelihood (Rodgers, Round, & Williams, 2012), and misrepresents the dynamism of in/formal economy interrelations, including the coexistence of partial formalisation with elements of ongoing informality (Williams & Lansky, 2013). The limited research on how microfinance facilitates the mobility of business from the informal to the formal economy (Bruton, Khavul, & Chavez, 2011) indicates that formalisation may occur to access loans (Gachet & Staehli, 2008, for small firms in Egypt), to access larger loans (Olapade, 2015, for India), or when the return on obtaining external funding exceeds the costs associated with formality (Amaral & Quintin, 2006; McKenzie & Sakho, 2010, p. 24), and that entrepreneurs’ motivation to formalise may change in response to constrained circumstances (Bruhn, 2013; De Castro, Khavul, & Bruton, 2014; De Mel, McKenzie, & Woodruff, 2013; Williams & Nadin, 2012; Williams & Shahid, 2016). We follow the lead of these emerging findings by seeing the relationship between microfinance use, enterprise growth, registration, formalisation, and entrepreneurial practice(s) as a matter for empirical investigation into the relationship between microfinance and small business development in situ.

This perspective requires research to grapple with the diversity of borrowers and the fluidity with which they move between different borrowing and/or entrepreneurial strategies in the face of constrained, and sometimes rapidly changing, circumstances. Whilst many studies have disaggregated borrowers, these have used theoretical-derived ‘researcher’ taxonomies which assume that the behaviour (and identity) of borrowers in these categories are ‘immutable’ (Jackson et al., 2019, p. 26). The differences between borrowers, entrepreneurs, their activities, the extent of their in/formality and their direction of travel, can be more meaningfully conceptualised in situ (Xheneti et al., 2019). Since enterprises and entrepreneurs adapt and change their activities as part of the evolving environment (Levie & Lichtenstein, 2010), formalisation in entrepreneurial practice needs to be understood as having ‘different causal pathways’ (Welch, Piekkari, Plakoyiannaki, & Paavilainen-Mäntymäki, 2011, p. 749). Such pathways admit room for failure, experimentation, iteration and reversals (Nguyen, Verreynne, & Steen, 2014; Sutter, Webb, Kistruck, Ketchen, & Ireland, 2017) and allow for a variety of linkages between microfinance use and borrower outcomes (Chen et al., 2017). Understanding these variations requires building analysis inductively from rich datasets rather than imposing, as much research does, predefined categorisations that give too much weight to prejudging how ‘entrepreneurial’ different borrowers are (Ramoglou et al., 2020).

Although country-level institutional arrangements¹ are acknowledged to mediate the performance of the microfinance sector (Ault & Spicer, 2014; Kummitt & Muñoz, 2017; Urbano, Aparicio, & Audretsch, 2019) and its impact on entrepreneurial development (Stenholm, Acs, & Wuebker, 2013), research on this front is overly focused on the tension between commercialisation, sustainability and growth of MFOs on one hand and the not-for-profits’ dependence on external donor funding for outreach, particularly to poor and ultra-poor clients, on the other (Hermes, Lensink, & Meesters, 2011; Olivares-Polanco, 2005; Quayes, 2015). Consequently, beyond this, relatively little is understood about how the commercialisation of MFOs enables (or constrains) entrepreneurial formalisation outcomes at the ‘bottom-of-the-pyramid’ in different contexts, although some research investigates the dynamism of microfinance and what this means for formalisation in specific places (Khavul et al., 2013; Lee & Hung, 2014). Improving our understanding of how microfinance is related to small business development in different contexts requires factoring in the institutional context in ways that go beyond perceived dualities between poverty alleviation and entrepreneurial practice(s) at the ‘bottom-of-the-pyramid’.

In sum, emerging literature argues for a substantial reorientation of research on microfinance and small business development towards a more exploratory inquiry that asks more open questions about how microfinance relates to entrepreneurial outcomes. Doing this requires employing methodologies that are more empirically-informed, attuned to the agency of everyday entrepreneurs, and more contextually-situated. This is what we do here.

3. The institutional context of microfinance and small business development in Kazakhstan

Small business development is regarded critical for countries that are transitioning from centrally-planned socialism to market socialism because large business is in its infancy, the economic infrastructure is still developing, and state employment and the state social provisioning is being dramatically retrenched (Davis, 2016). Small business development is widely regarded as having played an important role in Poland’s and Russia’s transitions (Bliss & Garratt, 2001; Davis, 2016; Tovstiga et al., 2004). The promotion of microfinance is seen as a strategy addressing the lack of capital at the point of transition, contributing to the emergence of economic infrastructure, and supporting the growth of new kinds of entrepreneurial activity. Microfinance is growing rapidly in transitional countries of Central Asia but relative to many developing countries, the sector is in its infancy: beginning later, from a small base, it lacks funding and has very low outreach, especially in rural areas, leaving many potential clients with poor cash-flows unserved.

In addition, centrally-planned socialism has also left many transitional economies with underdeveloped economic infrastructures and inefficient and often corrupt bureaucracies. Peng’s (2001) characterisation of entrepreneurs (as farmers, grey individuals, cadres or professionals) and their strategies (of prospecting, networking and boundary blurring) in transitional economies reveals the challenging and ambiguous nature of doing business in these contexts. Although contemporary entrepreneurialism has ‘roots in the underground economic activities’ that flourished in many centrally planned economies (Davis, 2016, p. 87), successful economic transition requires changes in people’s mindsets and symbolic structures around business and how resources are made available. These are uniquely challenging in former Soviet Union countries where ordinary citizens optimised their resilience by relying on strong informal networks, whilst deliberately constricting their links with the predatory state (Rose, 2000).

Kazakhstan is under-researched (Chen et al., 2017; Low, 2006). It experienced the highest degree of Russianisation of any of the post-Soviet nations (Davis, 2016; Low, 2006) and post-Independence (1990–1996) saw almost total economic collapse (GDP declined by 40%) and the sharp contraction of state sector employment (from 88.6 to 34.6%) (Davis, 2016; Verme, 2000). Nevertheless, Kazakhstan experienced a quick economic turn-around and this is reflected in its annual economic growth – averaging over 8 per cent for 2002–2007 (Davis, 2016).

Whilst much of Kazakhstan’s economic success was due to exploiting extractive natural resources (oil, gas and minerals), the growth in microenterprises played an important part too and suppressed even worse effects of transition, and the resulting recession, on unemployment and poverty (Davis, 2016). Kazakhstan’s early political support for entrepreneurial development mirrors that in Kyrgyzstan but differs from the other Central Asian economies which were slower to embrace a market economy and to develop enabling commercial law.

Microfinance has been a key component of the government’s national development strategy since the mid-1990s (United Nations Development Programme [UNDP], 2005). Almaty and Almatinskaya Districts account for a quarter of active entrepreneurs nationally (Statistical Agency of Kazakhstan [SAK], 2019) and Almaty city is reputed to have a strong ‘entrepreneurial spirit’ (UNDP, 2005, p. 41). Whilst the rate of business creation in Kazakhstan has been relatively high, it is dominated by small and medium enterprises (SMEs), which tend to have low productivity (Organisation of Economic Co-operation and Development [OECD], 2018, p. 15). Significant barriers for small businesses include poor infrastructure, a poorly educated workforce, an antiquated tax system, underdeveloped legal and financial institutions, excessive bureaucracy as well as inadequate access to finance (Davis, 2016; Luthans & Ibrayeva, 2006; OECD, 2018).

Kazakhstan’s legislation does not attempt to fully separate the formal and informal economy and distrustful entrepreneurs habitually hide their wealth and bribe government agents (Can, 2003; Davis, 2016; Suhir & Kovach, 2003). Few SMEs can borrow from banks because of their high interest rates and the extremely high value of collateral required: in 2013, only 19 per cent of SMEs had a bank loan or line of credit and only 6 per cent of SMEs used banks to finance investments (OECD, 2018). The 2008 economic crash had a severe impact ‘on the Kazakhstan banking system, which led to a credit crunch that has particularly affected SMEs’ (OECD, 2018, p. 88). With hindsight, 2012, the year when this study’s data collection was completed, sat at the nadir of the impact of this contraction of finance on SMEs.

In comparison to banks, MFOs in Kazakhstan offer individual loans with more flexible interest rates, less strict requirements for loan guarantees, the possibility of accessing further loans under simplified procedures and preferential conditions (after satisfactory borrower performance), as well as faster access to credit (UNDP, 2005, p. 45). The UNDP reported in 2005 that MFOs had extended the scope and spread of their financial services, but by this time donor funding was shrinking and MFO sustainability was undermined by laws prohibiting them from collecting savings and resistance to them registering as commercial banks (UNDP, 2005, p. 7). Outreach was poor and relatively few MFOs provided group lending with most offered guaranteed individual loans (UNDP, 2005, p. 42–43). In 2005, there were 60–70,000 microfinance clients borrowing from MFOs and commercial banks, but conservative estimates suggested that over 200,000 people needed access to basic financial services (UNDP, 2005, p. 39, 43). The persistently low average loan size in Almaty and Almatinskaya District (of less than 3,000 Kazakhstani Tenge [KZT]) suggests that MFOs largely meet a need for working capital rather than for investment purposes (SAK, 2019).

4. Methods

Following from our review of emerging research, we use a mixed-method approach to offer an in-depth empirical analysis of the relationship between microfinance and entrepreneurial practice in Almaty City and the surrounding Almatinskaya District in Kazakhstan. We regard the city and its surrounding district as a single case study ‘interesting in its own right’ (von Krogh, Rossi-Lamastra, & Haefliger, 2012, pp. 278–279) and seek a ‘contextualised explanation’ (Welch et al., 2011) of entrepreneurial practice(s) in relation to microfinance. Our methodology is designed to enable us to explore the relationship between MFOs and everyday entrepreneurial practice in Almaty. It is built around an innovative data-driven clustering approach to understand borrower diversity that offers advantages for contextualised analysis in comparison to more widely used theoretically-derived taxonomies of borrowers.

Data collection involved three phases. Phase 1 involved in-depth interviews with 6 key industry informants who were employees or founding members of AMFOK (the Association of MFOs of Kazakhstan) to obtain an overview of the sector and to broker access to local MFOs. Phase 2 comprised qualitative semi-structured interviews with 23 MFO loan evaluation officers and/or MFO owners/managers to investigate MFO interactions with entrepreneurs and generate data about MFO perspectives on their relationships with borrowers. Phase 3 consisted of a survey of 151 entrepreneurial borrowers identified by snowballing from current or former borrowers of the MFO respondents in phase 2.

The survey interviews were conducted face-to-face to enhance trust given Kazakhstan’s ‘closed context’ (Hisrich & Grachev, 1995; Koch, 2013). However, as other researchers found (Koch, 2013), recruitment was difficult with many unwilling to be interviewed and/or to refer the researcher to other entrepreneurs. The response rate for the survey was 18%, slightly higher than that reported (15%) in self-administered surveys of entrepreneurs in Kazakhstan (Seilov, 2015). Our snowball sample included borrowers using State MFOs even though we had not interviewed State MFOs in phase 2. Our borrower sample comprising 39 per cent informal entrepreneurs and 61 per cent formal entrepreneurs and is broadly representative of the business population (SAK, 2010).

We combined thematic analysis with cluster analysis to generate an empirically-derived taxonomy of borrowers. A detailed account of this process is provided in the Appendix A but, in brief, we analysed the themes emerging for qualitative analysis in phase 1 and 2 to develop a conceptually clustered matrix (Miles & Huberman, 2014). This was grouped around borrowers’ use of microfinance, their relationships with MFOs, and their patterns of formalisation and growth. This data structure underpinned our selection of clustering variables and the design of the borrower survey in phase 3. We analysed the borrower survey data using hierarchical cluster analysis to create the taxonomy and used three reliability tests to establish that cluster solutions were statistically different.

Our taxonomy of entrepreneurial borrowers serves as the backbone of our qualitative data presentation that adds flesh to numbers and shows the historical contingent circumstances that lead to certain entrepreneurial outcomes. In other words, we sought to make sense of our taxonomy by (re)embedding it within an analysis of our data from multiple sources. This design enabled an analytical focus on both process (qualitatively) and outcomes (quantitatively) of Kazakhstani microfinance lending and took account of the voice of lenders and borrowers. Our inductive identification of different clusters of entrepreneurial borrowers is a key methodological contribution. Accordingly, we reflect in the conclusion on its analytical value for microfinance and entrepreneurial research.

5. Findings

We begin by outlining the MFOs, before moving on to report our taxonomy of microfinance borrowers, their relationships with MFOs, their degrees of formalisation and their entrepreneurial pathways.

5.1. MFOs in Almaty and Almatinskaya Districts

The key informants understood the differences between MFOs on the basis of their ownership and distinguished between ‘foreign’ MFOs (including those established, funded and/or controlled by international development organisations), ‘private’ MFOs (owned by private Kazakhstanis, mostly financed by commercial bank loans) and ‘State’ MFOs (established by the State as and wholly funded by the State). These categories were commonly used in Almaty, so we have deployed them throughout.

Key informants argued that the majority of Kazakhstani MFOs, Private MFOs, were established to generate profit and mostly focus on individual collateralised loans. Whilst they felt that several MFOs were trying to maintain a social mission, these were largely the Foreign MFOs, which tended to focus on group loans. The third category, the State MFOs, are mostly focused on collateralised loans for registered agricultural entities. In contrast to the other types of providers, their application processes are longer, but their interest rates are lower (under 6–9% per annum).

The characteristics of the MFOs studied (see Table A1) confirm the key informants’ assessment and accords closely with secondary literature. The 12 Private MFOs in our sample were small-scale and few offered banking services to those without collateral. Whilst three Private MFOs in our sample had originally been set up as foreign-owned MFOs offering group loans, their transition to local private ownership signalled an end to group-based lending. Only two Private MFOs offered loans without collateral and only one of these offered group-based lending. In contrast, the Foreign MFOs, both of which were established in Almaty in 1997, constituted much more significant suppliers of microfinance. Both operated on a much larger scale, with large numbers of employees and many outlets, both offered group loans and un-collateralised loans.

5.2. Entrepreneurial borrowers in Almaty and Almatinskaya Districts

Our hierarchical cluster analysis was based on variations in borrowers’ entrepreneurial backgrounds, their enterprise characteristics and their relationships with MFOs. We identified six clusters: Necessity-Driven Entrepreneurs, Entrepreneurial Workers; Migrant Entrepreneurs; Entrepreneurial Professionals; Opportunity-Driven Entrepreneurs and Agricultural Business Entrepreneurs (see Table 1). We present these clusters by their degree of formality and the dynamism of their growth, for reasons of simplicity, but do not assume nor posit a specified order or a sequence of mobility across the spectrum.

Table 1. Entrepreneurial clusters in Kazakhstan

	Necessity-driven entrepreneurs	Entrepreneurial workers	Migrant entrepreneurs	Entrepreneurial professionals	Opportunity-driven entrepreneurs	Agricultural business entrepreneurs
Total	22	34	25	29	30	11
Females (%)	77.27%	94.11%	31.81%	79.31%	30%	18.18%
Mean age	40.27	41.85	42.04	45.59	43.6	50.82
Migrated (%)	13.63%	44.11%	60%	48.27%	20%	0%
Education (university degree)	0	20.58%	24%	75.86%	80%	45.45%
Blue-collar experience (%)	45.45%	47.05%	52%	13.79%	3.3%	0%
Professional experience (%)	31.81%	41.76%	16%	62%	46.66%	54.54%
Applied to banks (%)	40.9%	55.88%	52%	27.58%	36.66%	54.54%
Mean years of operation	5.76	7.9	9.8	12.03	8.97	11.64
Mean number of employees	4.23	1.76	4.16	4.76	14.93	9.45
Mean profit growth in (%) in 2008–2011	19.55%	51.21%	34.78%	1.55%	47.83%	32.27%
Mean turnover growth in (%) in 2008–2011	32.95%	70.38%	34.91%	−3.78%	55.83%	68.18
Family Business	63.63%	38.28%	76%	41.36%	36.67%	90.1%
Foreign MFO clients	63.6%	67.6%	12%	13.8%	3.3%	0
Private MFO clients	36.4%	32.4%	88%	86.2%	90%	9%
State MFO clients	0	0	0	0	6.7%	91%
Individual Collateralised Loans	18.18%	26.47%	84%	68.96%	100%	81.81%
Noticed MFOs’ influence on growth	77.2%	85.3%	68%	79.3%	76.7%	91%
Own livestock	27.27%	5.88%	40.00%	0.00%	30.00%	100%
Own apartment	27.27%	61.76%	24.00%	65.52%	56.67%	81.82%
Own a house	72.73%	38.24%	76.00%	27.59%	53.33%	100%
Own land	68.18%	23.53%	48.00%	27.59%	53.33%	100%
Own vehicle	40.91%	55.88%	80.00%	72.41%	93.33%	100%
Own other assets	13.64%	44.12%	40.00%	6.90%	20.00%	100%

The Necessity-Driven Entrepreneurs cluster strongly resembled the ‘survivalist’ category of Berger and Udell (1995) and over two-thirds were female. This cluster showed the lowest levels of accumulated assets, medium income and turnover growth rates, and they mostly ran non-registered informal enterprises.

The Entrepreneurial Workers cluster was also consistent with Berger and Udell’s (1995) survivalist entrepreneurs but had the highest overall proportion of entrepreneurs with previous experience as blue-collar workers (nearly half), was predominantly female (94%) and was distinguished by relatively higher growth patterns.

The Migrant Entrepreneurs cluster were mostly migrants (60%), were mostly clients of Private MFOs, and were more likely to be registered (nearly two-thirds), and male (over two-thirds). They were mostly growth-oriented entrepreneurs, with three-quarters being family businesses and were highly embedded in business networks (Berner, Gomez, & Knorringa, 2012).

The Entrepreneurial Professionals cluster mostly had professional experience (62%), were mostly female (79%) with businesses with low growth rates. In contrast to the previous three clusters, they were significantly more likely to be registered. They closely match the professional entrepreneurs identified in transitional economies by Peng (2001)

The Opportunity-Driven Entrepreneurs cluster was also comparable to Peng’s (2001) professionals’ cluster but demonstrated high growth, had accumulated substantial personal assets, had higher than average incomes from their mostly registered enterprises and was mostly male (70%).

The Agricultural Business Entrepreneurs cluster had registered agricultural businesses, which are established profitable entities. They were mostly middle-aged men, employed an average of 10 workers, and owned land, livestock, and vehicles. They primarily borrowed from State MFOs from whom they obtained loans on beneficial terms.

Different entrepreneurial clusters were largely served by different MFOs. Most Necessity-driven Entrepreneurs and Entrepreneurial Workers borrowed from Foreign MFOs and used their loans for day-to-day purposes. By contrast, Migrant Entrepreneurs, Entrepreneurial Professionals and Opportunity-Driven Entrepreneurs primarily borrowed from Private MFOs and were more likely to use the loans for business/investment expenditures. Nevertheless, all clusters included borrowers from two of the three types of MFO.

5.3. Entrepreneurial clusters and MFO – borrower relationships

The clusters were also differentiated by the nature of the MFO-borrower lending relationships (Canales & Greenberg, 2016) with regards to the loan distribution process, the character and duration of the relationship, and the flexibility to arrange payment delays (Table 2).

Table 2. MFO-borrower relationships

	Necessity-Driven Entrepreneurs		Entrepreneurial Workers		Migrant Entrepreneurs		Entrepreneurial Professionals		Opportunity-driven Entrepreneurs		Agricultural Business Entrepreneurs
Total	22		34		25		29		30		11
Relationship Type	Prof	Humane	Prof	Humane	Prof	Humane	Prof	Humane	Prof	Humane	Prof	Humane
Private MFO	4(36.7%)	4(36.7%)	3(20%)	8(42.1%)	6(75%)	16(94.11%)	8(72.7%)	17(94.4%)	10(100%)	17(85%)	0	1(33.3%)
Foreign MFO	7(63.3%)	7(63.3%)	12(80%)	11(57.9%)	2(25%)	1(5.8%)	3(27.3%)	1(5.6%)	0	1(5%)	0	0
State MFO	0	0	0	0	0	0	0	0	0	2(10%)	8(100%)	2(66.7%)
No of loans received (mean)	2.18	4.64	5.47	3.58	2.25	3.47	3.45	5.67	3.7	4.7	1.63	4
Loan application process (days)	7.81	11.09	8.93	5.21	4.38	4.06	3.9	2.92	7.1	3.6	48.375	12.33
Duration of loan (month)	12.12	8	8.87	9.16	9	36	9.64	18.11	11.4	14.5	56.25	14
Length of relationship with loan officer (years)	1.41	1.256	1.5	4.623	6.645	5.02	2.515	5.889	7.166	4.45	2.75	8.33
Length of relationship with MFO (months)	22.09	35.09	29.93	38.32	25	42.4	60.55	53.89	45.8	31.8	31	60
Number of entrepreneurs who switched MFOs	0	2(18.2%)	2(13.3%)	1(5.3%)	0	4(23.5%)	1(9.1%)	2(11.1%)	9(90%)	4(20%)	8(100%)	3(100%)
Delays payments	3(27.3%)	1(9.1%)	1(6.7%)	0	0	8(47.1%)	5(45.5%)	10(55.6%)	4(40%)	13(65%)	0	2(66.7%)
Loans for day-to-day purposes	6(54.5%)	6(54.5%)	15(100%)	17(89.5%)	3(37.5%)	9(52.9%)	7(64%)	15(65%)	3(30%)	6(30%)	3(37.5)	0
Loans for Business/investment expenditure	4(36.4%)	4(36.4%)	0	2(10.5%)	4(50%)	7(41.2%)	3(27%)	3(13%)	5(50%)	9(45%)	0	1(33.3%)
Loans to purchase livestock	1(9.1%)	1(9.1%)	0	0	0	1(5.9%)	0	0	1(10%)	0	3(37.5)	0
Loans for other purposes	0	0	0	0	1(12.5%)	0	1(9%)	5(22%)	1(10%)	2(10%)	2(25%)	0
Loans for consumption	0	0	0	0	0	0	0	0	0	3(15%)	1(12.5)	2(66.7%)

Note: Prof: Professional.

Most surveyed borrowers engaging with Foreign MFOs (64.4%) obtained group loans. Foreign MFOs maintained that structured monitoring ensured timely repayments and limited the influence of the borrowers on lending relationships. They described distance in the lending relationship as ‘professional’ and achieved it through arm’s-length and impersonal communication (structured standardisation), rather than as in commercial banking through frequency, planning and paperwork (bureaucratic administration). As one Foreign MFO manager explained: ‘Loans must be distributed as fast as possible, be as convenient and flexible as possible. It is the logic of microfinance … less paperwork’ (MFO 9). Whilst this ‘flexibility’ addressed the generic circumstances of marginal borrowers, it did not signal loan officer discretion or responsiveness to individual needs. Foreign MFOs described ‘non-professional’ lending relationships as dangerous: ‘a loan officer may become hostage of friendly relationships’ and consequently ‘he might not be able to refuse’ requests to borrow unrealistically large amounts (MFO 9).

By contrast, Private MFOs valued lending relationships described by the borrowers as being friendly, family-like, neighbourly and good. Private MFO officers and borrowers referred to these relationships as ‘chelovechno’ (Человечно) meaning ‘treating each other humanly’. A Private MFO owner (MFO 11) reported that they specifically target long-established borrowers with growth potential and talk to them ‘with respect … as a human [being]’ and ‘meet [their needs] half-way’ and another told us that ‘in relationship lending the most crucial thing is trust’ (MFO 13).

From the client perspective, borrowers engaging with Private MFOs primarily utilised individual collateralised loans (68.4%) and a more personalised lending relationship that was valuable for borrowers in terms of their ability to (re)negotiate favourable terms with their lender. For example, most migrant entrepreneurs (88%) borrowed from Private MFOs, with almost all (94%) defined their lending relationship as chelovechno: they were able to obtain long loans (averaging 36 months), maintained long relationships with their respective MFOs (on average 42.4 months) and almost half of them had delayed their loan repayments.

The Private MFOs also valued more personalised relationships with their clients because they enabled them to access soft information from their borrowers. Although Migrant Entrepreneurs, Entrepreneurial Professionals and Opportunity-Driven Entrepreneurs were less informal that the Necessity-Driven Entrepreneurs or Entrepreneurial Workers, some were unregistered and most were only partially formalised and were not able or willing to provide the evidence and guarantees that would make them bankable in the formal banking sector. One Private MFO loan officer (MFO 6) reported that ‘clients are secretive; withhold information about why they would like to borrow money’. The loan officer added that ‘[t]hey think that if they offer collateral, then only general information must be given’ but ‘that is not always possible for us to seize [the collateral offered]’. Thus, Private MFO clients were in a relatively empowered position in relation to their lenders and they exercised this power in their dealings with their lenders. Despite these generalised trends, both kinds of lending relationships (professionalised and personalised) existed with both kinds of MFOs (Foreign and Private).

The Agricultural Business Entrepreneurs cluster, largely served by State MFOs, mostly report impersonal and standardised (‘professional’) lending relationships and would not hesitate to switch between MFOs. Here, loan terms were up to five times longer than for loans from Private and Foreign MFOs and with considerably lower interest rates. Thus, this cluster received loans on a much more favourable basis than other microfinance clients because as registered and growing enterprises, with accumulated and ‘seizable’ assets, they were able to secure loans from State MFOs.

5.4. Entrepreneurial clusters and formalisation

The extent of registration of the clusters allows us to explore the degree and timing of formalisation of entrepreneurial borrowers and to examine how these were related to their growth and turnover (see Table 3).

Table 3. Registration and formalisation

	Necessity-driven entrepreneurs	Entrepreneurial workers	Migrant entrepreneurs	Entrepreneurial professionals	Opportunity-driven entrepreneurs	Agricultural business entrepreneurs
Total	22	34	25	29	30	11
Unregistered entrepreneurs
Unregistered (%)	21(95.45%)	22(64.7%)	7(28%)	6(20.68%)	1(3.3%)	0
Informal entrepreneurs that plan to register	13(61.9%)	14(64.6%)	6(85.7%)	3(50%)	1(50%)	0
Growth in profits of entrepreneurs planning to register	30.77%	35.36%	35%	0	50%	-
Growth in turnover of entrepreneurs planning to register	53.46%	78.71%	40%	16.67%	50%	-
Growth in profits of entrepreneurs not planning to register	−6.25%	26.88%	8%	7.5%	−20%	-
Growth in turnover of entrepreneurs not planning to register	−6.25%	26.88%	8%	7.5%	−20%	-
Registered entrepreneurs
	1	12	18	23	29	11
Registered (within corresponding cluster)	1(4.55%)	12(35.3%)	18(62%)	23(80.22%)	29(96.7%)	11(100%)
Registered first year of operation	1(100%)	2(16.7%)	5(27.7%)	5(21.7%)	4(17.3%)	1(9.1%)
Registered second to fifth year of operation		7(58.3%)	8(44.4%)	3(13%)	10(34.5%)	2(18.2%)
Registered their operation fifth to seventh year of operation		2(16.7%)	3(16.7%)	0	2(6.7%)	1(9.1%)
Registered after seventh year of operation		1(8.3%)	0	15(65.2%)	13(44.8%)	7(63.6%)
Registered entrepreneurs who do not declare all their income	1(100%)	6(50%)	4(22.2%)	12(52.1%)	8(27.6%)	3(27.3%)
Registered entrepreneurs who do not declare all their employees	1(100%)	2(16.6%)	4(22.2%)	9(39.1%)	7(24.1%)	3(27.3%)
Growth in turnover in 2008–2011	80%	85.92%	28.61%	−8.26%	55.13%	68.18%
Growth in profits in 2008–2011	80%	89.67%	29.44%	6.5%	50.18%	32.37%

Almost no Necessity-driven Entrepreneurs had registered their enterprise, but those who were planning to register, on average, experienced a higher growth in profits and turnover than those who were not eager to register. However, a third did not feel that microfinance had an impact on their growth. The sole registered borrower in this cluster reported very healthy growth but did not declare all their income or all their employees.

Although most Entrepreneurial Workers were somewhat more formalised than the Necessity-driven Entrepreneurs cluster and over half of those registered did not declare all their income and/or employees. Here, registration and the intention to register were again associated with higher growth patterns. Whilst some registered in their first year of operation, most registered in their second year of operation and others in later years indicating that they choose when to register their activities rather than being compelled to (either by MFOs or by the dynamism of their businesses’ growth). Indeed, a Foreign MFO officer told us: ‘when [borrowers] themselves see [registration] as necessary, [then they will] obtain it … We say that is preferable, but not mandatory’ (MFO 9).

In contrast, the majority of Migrant Entrepreneurs and Entrepreneurial Professionals had registered their business, and the majority of those who had registered reported all their income and/or employees. Not only were Migrant Entrepreneurs more likely to have registered in the early years of their operation, those who were not yet registered were more likely to plan to register than any other cluster. However, in a reversal of the patterns above, the non-registered Migrant Entrepreneurs with plans to formalise reported higher growth rates than the registered ones. Tentative explanations include: migrant entrepreneurs were mostly men, who tend to be more formal (Khamis, 2014); migrants in Kazakhstan tend to be more self-reliant and entrepreneurial, whilst locals expect governmental support (Sancak, 2007; Sancak & Finke, 2005); state surveillance of their alien status is more insistent upon registration; and/or, lacking informal local support networks, migrant entrepreneurs perceived formalisation as an opportunity to grow.

Whilst the majority of Entrepreneurial Professionals were also registered, most only registered after their 7th year of operation and only half of those unregistered were planning to obtain registration. Moreover, the average turnover of those who were registered was negative and lower than those who were unregistered and this cluster was the least likely of registered entrepreneurial borrowers to declare their income and/or employees fully to the tax authorities. These findings suggest that formalisation, especially early and full formalisation, may be problematic for Entrepreneurial Professionals. Whilst local Kazakh with higher education expected governmental support, they have high levels of disillusionment with government (Sancak & Finke, 2005): this may motivate them to avoid full formalisation, postpone registration, and/or find ways to keep some degree of informality.

At the more formalised end of the spectrum, sat Opportunity-Driven Entrepreneurs and Agricultural Business Entrepreneurs. All but one of the Opportunity-driven entrepreneurs were registered, and all report high average growth. Even though all Agricultural Business Entrepreneurs were registered, owned assets and could secure State MFO loans, they did not embrace their formal status early on or unconditionally. A majority declared all their income and employees, but a quarter did not, and many only registered after operating for 7 or more years.

As with lending relationships, these generalised trends coexisted with significant variations in formalisation, and its relationship with growth, within as well as between clusters underlining the existence of multiple and historically contingent entrepreneurial pathways.

5.5. Fuzzy clusters and multiple pathways

Whilst differentiating six clusters of borrowers, we note a degree of ‘fuzziness’ (Khavul et al., 2013) between and within the clusters. Whilst belonging to distinguishable clusters, individual borrowers’ pathways varied because they exerted agency in their everyday entrepreneurship and because their interactions were nested with the wider changing business context. Moreover, the specific characteristics of marginalised entrepreneurial borrowing here – notably the undersupply of microcredit, the preference for close informal relationships and suspicion about formal bureaucracy, and the possibility of partial formalisation – militated towards uncertainty in relation to formalisation and its supposed ‘rewards’.

We found that strong growth did not necessarily infer registration, either immediately or over the longer term, nor did weak growth necessarily preclude it. This mismatch between formalisation and profitability was acknowledged by MFOs and fed into how Private MFOs approach building personalised relationships with borrowers in this context. For instance, a Private MFO loan officer told us: ‘we have a big client who sells wood (4 million turnover) and owns a big part of this area, officially she doesn’t show anything, but everyone knows her’ (MFO 7). The soft information that Private MFOs accessed protects them against taking on risky investments, allows them to enact productive opportunities in the informal sector, and effectively links the micro entrepreneur with capital flows. As another Private MFO owner noted: ‘if [borrowers] are not able to prove [their income] on paper but can prove it in real life [by working]. Why not? We don’t need proof, we do it [proof-checking] ourselves’ (MFO 6).

Our inference that formalisation did not matter for Foreign or Private MFOs was supported by loan officers who reported that they ‘have no right to intervene’ (MFO 6) in their clients’ decisions about formalisation. Indeed, it may not be in their interests to promote client formalisation, as Private MFO loan officers recognised: ‘as soon as [the borrower] becomes a respectable citizen, [that is] a legal entity, he is already not our client’ (MFO 6) as he becomes eligible to apply for State MFO and/or commercial bank loans.

Whilst the entrepreneurial borrowers surveyed did not go through predictable stage-driven pathways towards defined endpoints, they were robustly clustered around background characteristics (including gender), enterprise characteristics, and patterns of microfinance use. Moreover, many at some point did decide to ‘leave the shadow economy, from being clearly informal where they do not pay any taxes, [and move] into a grey area where they start to partially pay taxes’ (Private MFO 11). Whilst the ‘grey-ness’ of partial formalisation appeared to generate or promise room for growth for some clusters of entrepreneurial borrowers, such as Opportunity-driven Entrepreneurs, it appeared to be problematic for others, such as the Entrepreneurial Professionals. The existence of these varied experiences and multiple pathways suggests that entrepreneurial borrowers, to varying extents, decide whether, when and how to formalise. Significantly though, borrower agency over formalisation was greater or lesser in some clusters than in others and was firmly embedded in the wider institutional context.

6. Discussion and conclusion

Whilst similarities can be drawn between some of our clusters and theoretically-derived categorisation of borrowers, such as the ‘survivalist’ (Berger & Udell, 1995) or ‘professionals’ (Peng, 2001) categories, our empirically-derived methodology offers a finer grained analysis that yields important insights. Collapsing Necessity-Driven Entrepreneurs and Entrepreneurial Workers into a single ‘survivalist’ category, or collapsing Entrepreneurial Professionals cluster and the Opportunity-driven Entrepreneurs cluster into a ‘professionals’ cluster, would render invisible the greater significance of microfinance for small business development for Entrepreneurial Workers and Opportunity-Driven respectively. These findings question the value of theoretically-derived categories which make assumptions about who uses microfinance to service everyday needs and which borrowers have the potential to develop their enterprises through microfinance use. In contrast, our innovative methodology offers an alternative way forward for researchers seeking to go beyond these assumptions to investigate the empirical reality of ‘everyday’ entrepreneurship of microfinance borrowers in specific contexts (Xheneti et al., 2019).

Our findings suggest that the stages-model assumption of formalisation (Levie & Lichtenstein, 2010) has little traction in transitional settings. Although some unregistered borrowers who made extensive use of microfinance exhibited higher growth rates and expressed a propensity for future registration, others with extensive growth experience and substantial income never formalised their enterprises. Moreover, similar degrees of formalisation had different implications for microfinance use for different borrowers. For instance, formalisation ensured Agricultural Business Entrepreneurs could access loans from State MFOs, whilst formalisation for Migrant Entrepreneurs appeared to legitimise their entry to doing business in Kazakhstan thus enabling them to establish lending relationships with Private MFOs. For most entrepreneurial clusters, informality or even partial informality appeared to be an important way of addressing uncertainty and mobilising resources in the operation of enterprises in this institutional context.

Whilst standardised lending relationships are considered a good practice for MFOs (Akula, 2008; Chakrabarty & Bass, 2014), our findings suggest that the personalised chelovechno relationships are more effective in this transitional context. As in other transitional settings, entrepreneurs in Almaty and Almatinskaya districts were reluctant to share business information, often partially or entirely hiding their businesses from registration or formal declaration, whilst MFO lenders found it hard or impossible to seize assets in cases of default. Personalised relationships were significant for both Private MFO providers (to access soft information to guard against unsound loans and enact good investment opportunities) and MFO borrowers (allowing them to negotiate more favourable and more flexible loan terms). So although individual loans were described as ‘collateralised’, in reality, the soundness of lending was verified and secured through chelovechno relationships. This kind of relationship lending is not textbook but, given the context in which they operate, ensured that Private MFOs can repay the formal bank loans which they take to finance their microlending.

Our study questions three theoretical assumptions about how microfinance drives small business development in transitional contexts: first, the assumption that microfinance is (or should be) primarily ‘for’ poverty alleviation or small business development; second, the assumption that microfinance drives small business development through formalisation; and third, the assumption that microfinance drives formalisation by inculcating standardised lending relationships.

Firstly, rather than fuelling debate over the relative emphasis on promoting ‘sustainability versus outreach’ in microfinance (Ault & Spicer, 2014; Shahriar, Schwarz, & Newman, 2016), our study demonstrates the complementarity of a range of MFOs for small business development in transitional contexts like Kazakhstan. The simplistic expectation of dualistic perspectives, that the clients of Outreach MFOs are ‘poor’ and primarily use their loans for everyday purposes whilst those using Commercial MFOs are ‘would-be’ entrepreneurs and primarily invest their loans in developing their businesses, were not borne out in Almaty and Almatinskaya districts. Here a substantial proportion of ‘poor’ borrowers use Commercial MFOs and invest in business development, and a significant minority of ‘would-be entrepreneurs’ borrow from Outreach MFOs for everyday expenses. In other words, both Private and Foreign MFOs in Almaty played an important role in enabling ‘unbankable’ borrowers to deal with ongoing and unexpected livelihood needs at the same time as, and by, trying to build up small businesses. This complex reality underlines the importance of a diversity of microlenders, and the fungibility of their loans, for ‘everyday entrepreneurs’ who face rapidly changing opportunities and contingencies in the informal sector in transitional contexts like Kazakhstan.

Clarity over the intrinsic inter-relationship between the poverty alleviation and small business development at the bottom-of-the-pyramid in Kazakhstan makes room for a policy focus on undersupply of microfinance. The problem in this transitional context is not the commercialisation of microfinance (Hoque et al., 2011): here, for-profit MFOs were important economic agents of development who engage closely and personally with the social worlds of their clients (Beck et al., 2018; Biggart & Delbridge, 2004). Rather the problem is that demand for microfinance outstrips supply across the board: both Foreign MFOs (that service needs for group lending) and Private MFOs (that service needs for individual lending) were constrained in terms of the finance – the former by dwindling donor financing, the latter by resistance to larger commercial MFOs registering as banks, and both by constraints on holding savings. Enabling diverse MFOs to hold savings has considerable potential to grow the microfinance sector and extend its contributions to fostering small business growth and economic development in Kazakhstan.

Secondly, our research questions the assumption that extensive microfinance use promotes formalisation (Xheneti et al., 2019) through business growth (Chen et al., 2017; Chliova et al., 2015) in transitional settings. In Almaty and Almatinskaya districts, neither Foreign nor Private MFOs promoted or required formalisation, they only required repayment and the extent of microfinance use by clients, and the growth of their enterprises, were delinked from formalisation. Here, all microfinance borrowers exhibited ‘shades of grey’ (De Castro et al., 2014) combining informal and formal activities (Xheneti et al., 2019) depending on the changing requirements of their everyday entrepreneurial needs (Beck et al., 2018; McKelvie & Wiklund, 2010; Sutter et al., 2017; Welter et al., 2017).

Consequently, formalisation rates are unlikely to be a reliable proxy for small business development or an effective measure of how far microfinance is supporting small business development in transitional contexts where a vibrant informal sector maintains a distrust of the state as predatorial/corrupt. Furthermore, policy encouraging MFOs to push clients towards formalisation (International Labour Office (ILO), 2021) would be counterproductive in these settings for both MFOs and for small business development. More contextually sensitive and multi-dimensional assessments (including social metrics) of the relationship between microfinance and small business development are needed to improve both impact evaluation and inform policy direction in transitional settings (Behr, Entzian, & Guettler, 2011; Dowla, 2006).

Thirdly, we question the assumption that microfinance inculcates ‘unbankable’ borrowers into standardised borrower-lender relationships (Akula, 2008; Chakrabarty & Bass, 2014) – thus paving the way for them to graduate to the formal banking sector – in transitional settings. In Almaty, the chelovechno relationships preferred by Private MFOs and their clients did not mimic universalised formal banking relationships. Rather, they were hybridised: relying on soft not hard information, on trust not (un)seizable collateral, on lending over time rather than repayment in time, and on culturally-appropriate personalised relations rather than impersonal professional relations. These relationships were carefully attuned to enabling the lender and ‘unbankable’ borrower to do business with one another in this high-risk setting and worked well to establish reliable repayment relationships in which lenders and clients were both invested (Behr et al., 2011).

Our analysis suggests that microfinance sectors operating in other transitional settings may (need to) evolve hybridised borrower-lending relationship in order to do business in a way that deals with the risks that these settings raise for operating microlending. Following from this, a policy focus on standardising or professionalising micro-lending relationships risks ignoring the potential value of hybridised lending relations that have evolved in situ and may be inimical to the interests of either the microfinance sector or small business development in transitional settings.

To conclude, our study confirms the importance of context-sensitive research on the relationship between microfinance and everyday entrepreneurship in specific institutional settings. We have contributed an innovative methodological approach to this endeavour and our rich empirical findings from Almaty and Almatinskaya districts demonstrate that the relationship between microfinance and everyday entrepreneurs in transitional settings needs rethinking. To this end, further research to investigate the realities of microfinance’s quotidian operations and its interactions with clients (Beck et al., 2018) is urgently needed in transitional economies.

Acknowledgements

The authors thank Haya Al-Dajani for acting as a scientific advisor during the data collection and analysis of this study and for helpful comments on earlier drafts of this manuscript. Special thanks also go to the Editor Richard Palmer-Jones and two anonymous reviewers for their challenging feedback that has made us return to our analysis and rework the argument with much improved insight. Both the study’s quantitative MFO borrower survey database and related cluster analysis (with annotation to facilitate replication) is available upon reasonable request. A supplementary material with the data structure of themes used in the study’s qualitative analysis of in-depth interviews with key industry informants and MFO’s loan officers is also available for this article which can be accessed via the online version of the journal at: …

Disclosure statement

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

Supplementary material

Supplementary Materials are available for this article which can be accessed via the online version of this journal available at https://doi.org/10.1080/00220388.2021.1956472.

Notes

1. Salient country-level institutional arrangements for the relationship between microfinance and small business development comprise the wider business environment, including the nature of the state (e.g. its fragility), the inter-related in/formality of the economy, historically-specific political and social dynamics that have a bearing on the contemporary character of economic life, and the way in which these local factors are situated in relation to related global dynamics.

Appendix A.

Respondent selection

The selection criteria for MFOs in phase 2 were that MFOs were all members of AMFOK operating in Almaty and Almatinskaya and that they had more than 5 borrowers. 15 MFOs were contacted and up to two interviews were requested per MFO. In total, 23 interviews were held with 13 MFO owners/managers and 10 MFO loan officers who directly communicate with borrowers (see Table A1 below).

The entrepreneurial borrowers surveyed in phase 3 were identified using snowball sampling. We began in Phase 2 by asking each of the 23 MFO loan evaluation officer/owner to identify at least 5 current or former borrowers. When surveying these initial respondents, we asked them to refer other borrowers we could invite to be surveyed: those who agreed, were in turn asked to refer other borrowers for us to approach.

Cluster analysis

We used a conceptually clustered matrix (Miles and Huberman, 2014) method to develop a data structure from the themes that emerged from our qualitative analysis in phase 1 and phase 2. We used an iterative process of internal ‘peer debriefing’ to challenge the emerging themes and sub-themes (grouped through a process of open and axial coding), address potential biases and increase our understanding of the social and financial preferences of MFOs’ decision-making. This data structure (see Figures S1(a) and S2(a) in supplementary online appendix) underpinned our selection of our clustering variables.

Hierarchical cluster analysis (Ward Linkage Method and Squared Euclidian Similarity Distance) was subsequently employed to analyse the borrower survey data and create the taxonomy. Cluster analysis is targeted at maximisation of similarity between the cases within the groups and simultaneously minimising similarity between groups (Burns & Burns, 2009). Since the assessment of cluster reliability is inextricably interrelated with the assessment of cluster stability (Hair, Black, Babin, & Anderson, 2010; Mooi & Sarstedt, 2011), we used three reliability tests on the cluster solutions obtained: changing the order of cases to ensure stability, sample spilt and usage of different linkage measurements. Finally, ANOVA (Analysis of Variance) was used to establish that the clusters were statistically different in respect of our clustering variables and Effect Size Testing indicated how strongly the heterogeneous clusters were differentiated (ω) (Burns & Burns, 2009).

In our survey (phase 3) we interviewed entrepreneurial borrowers who were clients of all three types of MFO. Our selection of clustering variables used were divided into three distinctive groups: entrepreneurial background, enterprise characteristics and MFO relationship. This clustering has proved robust according to our three reliability tests and ANOVA testing showed that the means of the clustering variables differed in a statistically significant way across the six clusters (see Table A2 below). The most differentiating variables were: (a) whether a respondent had an unregistered enterprise, a registered agricultural enterprise or a joint stock enterprise; (b) whether a respondent received a group or individual collateralised loan; and (c) whether the respondent was a borrower of a Private, Foreign or State MFO. Whilst respondent’s gender and low education level were also important clustering variables, the length of time with current MFO has the weakest power of differentiation among the significant variables.

Table A1. Characteristics of MFOs studied in Phase 2

MFO	Type	Location of Clients	Year est.	Employees	No of outlets	Group loans	Loans without collateral
MFO 1.	Private	Almaty	2011	3	1	N	N
MFO 2.	Foreign	Almaty and Almatinskaya	1997	>100	24	Y	Y
MFO 3.	Private*	Almaty and Almatinskaya	1998	10	1	N*	N
MFO 4.	Private	Almaty and Almatinskaya	2009	11–12	1	N	N
MFO 5.	Private*	Almatinskaya	1994	3–4	1	N*	N
MFO 6.	Private	Almaty	2011	11	1	N	N
MFO 7.	Private	Almaty and Almatinskaya	2009	6	1	N	N
MFO 8.	Private	Almaty	2009	3–4	1	N	Y
MFO 9.	Foreign	Almaty and Almatinskaya	1997	>600	No data	Y	Y
MFO 10.	Private*	Almaty and Almatinskaya	2003	5–6	2	N*	N
MFO 11.	Private	Almaty and Almatinskaya	2004	10	4–5	N*	N
MFO 12.	Private	Almaty and Almatinskaya	1999	8	No data	Y	Y
MFO 13.	Private	Almaty and Almatinskaya	2003	7	1	N	N
MFO 14.	Private	Almaty	2004	3	1	N	N

Note: *Originally foreign-owned and originally offering group loans.

Table A2. Clustering variables and analysis of variance (ANOVA)

Clustering Variables	Possible Answers	F(5, 145) (ω)
Entrepreneurial Background
Age (Year); Gender; Education (Secondary; College; University; Postgraduate)	Binary (0 = no, 1 = yes)	2.38* (0.21); 15.90** (0.57); 15.39** (0.57); 1.49 (0.13); 12.33** (0.52); 2.58 (0.52)
Migration (from different urban area; from different rural area; from abroad)	Binary (0 = no, 1 = yes)	4.87** (0.34); 1.78 (0.16); 4.67** (0.33)
Parental occupational experience (professional; white collar; blue-collar)	Binary (0 = no, 1 = yes)	5.07** (0.34); 1.30 (0.10); 7.20** (0.41)
Own occupational experience (professional; white-collar; blue-collar)	Binary (0 = no, 1 = yes)	2.89* (0.24); 6.85** (0.40); 6.15** (0.38)
Enterprise Characteristics
Years of operation (1 to 25)	Scale	4.22** (0.31);
Type of organisation (Family business; partnership; working with spouse)	Binary (0 = no, 1 = yes)	4.69** (0.33); 6.90** (0.40); 3.00* (0.25)
Employment (1 to 110 employees); % Turnover growth (2008–11); % Profit growth (2008–11)	Scale	5.18** (0.35); 2.56* (0.22); 2.24^+ (0.20);
Vehicle owned; apartment owned; house owned	Binary (0 = no, 1 = yes)	4.93** (0.34); 4.56** (0.32); 9.64** (0.47)
Business registration type (Private entrepreneur; private entrepreneur with simplified declaration; private entrepreneurs with single-payment coupons; private entrepreneur on basis of patent; joint stock entity; business partnership; is not registered but working with single-payment coupons; is not registered; agricultural entity)	Binary (0 = no, 1 = yes)	2.99* (0.25); 8.94** (0.46); 2.28^+ (0.20); 4.17** (0.31); 34.44**(0.72); 2.10 (0.19); 15.38** (0.57); 103.67** (0.88); 107.43** (0.88)
MFO Relationship Characteristics
Client of (Private MFO; Foreign MFO; State MFO)	Binary (0 = no, 1 = yes)	17.29** (0.59); 17.77** (0.60); 55.43** (0.80)
Month with current MFO (1 to 168 months)	Scale	2.29* (0.20)
Number of Successful MFO loans (1 to 20)	Scale	1.44 (0.12)
Expenditure Use Type (Investment; Day-to-day purpose; consumption; business; buying livestock)	Binary (0 = no, 1 = yes)	6.05** (0.38); 8.33** (0.44); 1.54 (0.13); 4.37** (0.32); 9.99** (0.48)
Duration of loan (1 to 84 months)	Scale	12.06** (0.52)
Loan Type (group loan, uncollateralised; credit line, collateralised; individual, collateralised loan; individual, uncollateralised loan)	Binary (0 = no, 1 = yes)	30.83** (0.70); 4.49** (0.32); 16.18** (0.58); 3.91* (0.30)
MFO – Borrower Relationship (professional; neighbour; humane, friendship)	Binary (0 = no, 1 = yes)	1.42 (0.12); 3.00* (0.25); 1.49 (0.13); 1.30 (0.10)
Number of successful bank loans (0 to 20)	Scale	3.45^+ (0.27)

Note: ⁺p < .10; *p < .05; **p < .01.