Introduction

Small businesses have a significant aggregate economic impact (Neumark et al. 2011; Odell 2010; Wennekers and Thurik 1999). However, a key obstacle for early-stage entrepreneurs and small business owners is securing adequate financing to start and maintain their businesses (Ebbers and Wijnberg 2012; Mata 1994). Traditional sources of venture funding have included ‘in-crowd’ lenders (Polzin et al. 2018) such as self/family/friends, bank loans, angels, and venture capital (Burgelman and Hitt 2007; De Clercq et al. 2006). Yet, many entrepreneurs being small and unestablished struggle to attain financing from these traditional sources and must instead pursue alternative ‘out-crowd’ (Polzin et al. 2018) financing options (Desa and Basu 2013), such as lending-based crowdfunding.

While crowdfunding is rapidly on the rise (Bruton et al. 2015; Cortese 2013) and enjoying government support (JOBS Act 2012; Stemler 2013) to encourage capital raising for new ventures, scholars still know very little about the dynamics of crowdfunding (Short et al. 2017) and its implications for the business ethics literature (André et al. 2017). Indeed, analyses of particular crowdfunding efforts are few, and “whether crowdfunding efforts reinforce or contradict existing theories about how ventures raise capital and achieve success” (Mollick 2014, p. 1) remains inconclusive. Furthermore, while research has investigated motives driving entrepreneurs’ decision to use crowdfunding as a source of capital (Belleflamme et al. 2013; Lehner 2013), our understanding is still limited about the motives that influence lenders to fund such endeavors (André et al. 2017; Bretschneider and Leimeister 2017; Cholakova and Clarysse 2015). This study thus seeks to fill this gap by expanding our understanding of the motives that drive a specific type of lenders, lending-based prosocial crowdfunders, to fund entrepreneurs seeking capital through crowdfunding. Specifically, we explore whether lenders in this context follow traditional value-optimizing thinking by investing strategically on signals of quality, or whether they are more ethically driven and follow altruistic motives.

We build from prior research (e.g., Scholtens 2009), by using a social responsibility lens, to examine whether lenders follow a strategic or altruistic motive when lending online. We use Hemingway and Maclagan (2004)’s proposed framework for analyzing social responsibility, which highlights the individual as a locus of responsibility. Under this perspective, we study whether crowdfunders make prosocial lending decisions based on signals of quality (i.e., strategic lending) or based on idealistic characteristics which appeal to them personally (i.e., altruistic lending). In doing so, we are able to assess the impact of both strategic and altruistic motives in a global crowdfunding setting. Our empirical model follows prior work which conceptualizes and tests financial versus non-financial motives (for example, see Nielsen and Riddle 2010). We also build on recent work by André et al. (2017), which explores the relationship between altruism and self-interest in another type of crowdfunding, rewards-based.

To test our hypotheses, we examine a large sample of loans made on Kiva between 2008 and 2013. Kiva is recognized as the largest lending-based prosocial crowdfunding platform (Needleman 2010), operating in 69 countries (Allison et al. 2015). Kiva strives to connect vulnerable populations of entrepreneurs with global lenders through a network of local micro-lending institutions called field partners. Entrepreneurs’ profiles on Kiva are used to divulge signals about themselves and their venture (Courtney et al. 2017; Moss et al. 2015; Smith et al. 2017). Our results indicate that lenders on Kiva favor lending decisions based on signals of loan quality. Furthermore, we find that lenders may actually find altruistic characteristics as detrimental. This latter finding is noteworthy since it runs counter to prior work on lending-based prosocial crowdfunding (e.g., Allison et al. 2013, 2015), while also having important ethical implications. Notably, though, we also find that projects that are high on both financial and social appeal receive the highest average amount of funding.

This study makes three main contributions. First, this study makes an important contribution to the business ethics literature. Indeed, we explore the concept of social responsibility—acting responsibly or being ethical—from the perspective of the individual. Prior research indicates that individuals driven by ethical purposes may alter their altruistic behavior when the effects on society seem distant or uncertain (Vitell 2015). We argue that prosocial lending-based crowdfunding allows lenders to combine moral and social values with financial objectives, consistent with research on ethical individuals in other contexts (Hill et al. 2007). Nonetheless, whether the primary motive to lend is based on strategic or altruistic indicators, and the extent of that drive (i.e., intensity), is still misunderstood. We find that lenders on Kiva are primarily driven by strategic motives. However, an altruistic motive led by ‘social appeal’ of the entrepreneur’s profile does exist and seems to act as the driver for the intensity of the lender’s decision, as indicated by the positive correlation between social appeal and the average amount of funding the entrepreneur is able to accrue. With social responsibility accounting for almost 3 trillion dollars in investments in the current financial market (SIF 2009), and one in every eight dollars in financial markets being invested using a social responsibility lens (Laufer 2003), this study is both timely and salient.

Second, while the field has widely documented the financial constraints faced by young ventures, “the academic literature on most of the recent trends in entrepreneurial finance is still in its infancy” (Block et al. 2018, p. 2). As such, we address calls to extend theory as it pertains to crowdfunding (McKenny et al. 2017), while also understanding the motives that drive lenders in this new funding context (Bretschneider and Leimeister 2017). Specifically, our study takes a social responsibility lens (e.g., Hemingway and Maclagan 2004) to add to the limited knowledge on motives that influence prosocial micro-lending decisions (e.g., Allison et al. 2015; Bretschneider and Leimeister 2017; Scarlata and Alemany 2010). Our findings offer important insights into theoretical boundary conditions of social responsibility beyond traditional financial linkages (Scholtens 2006).

Finally, our study also contributes to the crowdfunding literature by examining an understudied, but significant type of crowdfunding—lending-based prosocial crowdfunding. Whereas much of the crowdfunding literature has focused on popular rewards-based crowdfunding sites such as Kickstarter (e.g., Calic and Mosakowski 2016; Colombo et al. 2015; Kuppuswamy and Bayus 2017; Mollick and Nanda 2015; Parhankangas and Renko 2017), few studies have examined other platforms servicing very different entrepreneurs and lenders. Research is particularly sparse regarding prosocial platforms (e.g., Allison et al. 2013, 2015; Burtch et al. 2014) and online peer-to-peer (P2P) lending sites (Iyer et al. 2015; Lin et al. 2013). Our findings are not only important for researchers trying to understand the theoretical underpinnings of lending-based prosocial crowdfunding, but also for practitioners around the world whose economic impact and opportunities are growing exponentially (Short et al. 2017).

Theory and Hypotheses Development

Crowdfunding and Prosocial Crowdfunding

Crowdfunding occupies an overlapping middle ground between micro-finance (Morduch 1999) and crowdsourcing (Poetz and Schreier 2012), making crowdfunding its own distinct research area (Ghezzi et al. 2017). Broadly defined, crowdfunding refers to online fundraising accessible to the general public, and which in general, seeks to provide financial resources to specific entrepreneurial projects or causes (see Burkett 2011; Lin and Viswanathan 2015; Mollick 2014; see; Schwienbacher and Larralde 2010 for other complementary definitions). Mollick and Robb (2016) further call crowdfunding the democratization of innovation and financing, which with the advent of the Internet offers increasingly “speeding and scaling opportunities for early-stage financing” (Vismara 2016, p. 580).

Compared to more traditional types of entrepreneurial funding (e.g., angel, venture capital, etc.), crowdfunding offers several advantages for entrepreneurs, including easily accessible funding, financial risk avoidance, the ability to overcome the disadvantage of the liability of newness, and access to a global target audience (Martinez-Cañas et al. 2012). It comprises an ever-growing segment of funding available to entrepreneurs, and accounts for a significant and global economic impact (Odell 2010). The benefits accrued from crowdfunding platforms are increasingly a driving force behind its popularity and growth in both practice and research. Despite the economic significance of this growing phenomenon, however, researchers have only just begun to examine and understand the intricacies that encompass the dynamics of crowdfunding (Short et al. 2017). Fortunately, increases in users’ online footprints are making the study of this expanding phenomenon both relevant and accessible.

With ever-expanding technological reach, crowdfunding participation has grown in terms of both entrepreneurs and lenders (Armendáriz and Morduch 2010). Platforms operate by connecting entrepreneurs with lenders whose funding is dispersed through equity offerings, debt, rewards, or donations (Mollick 2014). Platforms vary in their goals and operating models. For example, donation-based crowdfunding platforms such as Gofundme, rely on gifts with nothing expected in return. On the opposite end of the crowdfunding spectrum, depicted in Fig. 1, are rewards-based platforms such as Kickstarter, where lenders receive a non-monetary reward, such as free copies of finished art projects. Between these two platform types, entrepreneurs operate in a more traditional way by offering equity in the firm to accredited lenders (e.g., CircleUp), or by paying back debt (e.g., Funding Circle).

Fig. 1
figure 1

Spectrum of the various types of crowdfunding

Full size image

Yet, crowdfunding platforms such as Kiva, connect small entrepreneurs with lenders from all over the world who are willing to provide funding in the form of micro loans. In this model, the online platform fosters the connection between crowdfunding lenders and entrepreneurs and safeguards that relationship by providing oversight via local financial institutions (i.e., field partners) in each country. Micro-lending of this nature has been proposed as a way to stimulate economic development in poor areas of the world (Armendáriz and Morduch 2010; Bardy et al. 2012; Hollis and Sweetman 1998; Rankin 2001). The combination of the sheer number of loans, funds disbursed, number of participants, and the enabling technological media equates to a vast and multifaceted area of research, with global ethical implications and underpinnings (Snoy 1989).

Thus, when studying crowdfunding activities, it is important to consider which type of crowdfunding is being investigated: donation-based, lending-based, equity-based, or rewards-based crowdfunding (Ahlers et al. 2015; Bretschneider and Leimeister 2017). Even within one type of crowdfunding such as lending-based crowdfunding, there can be platforms with differing purposes. In particular, platforms may promote, to a greater or lesser extent, a prosocial agenda (Burtch et al. 2014). For example, Kiva markets itself as a prosocial lending-based platform which offers lenders a chance to aid those less fortunate with a loan which reaps the investors no interest. This type of platform is consistent with Allison et al. (2015), which suggest that some crowdfunding lenders tend to consider not only extrinsic factors, but also intrinsic factors such as prosocial motivation, when making lending decisions. In contrast, platforms such as LendingClub and Prosper offer investors a chance to lend with an emphasis on the accumulation of financial reward via interest rates. This type of platform follows a more traditional financial motivation of wealth creation for investors.

Social Responsibility and Ethical Lending in Prosocial Crowdfunding

From its very beginning, corporate social responsibility (CSR) was considered as being at the intersection of social-service and profit-making (Berle 1931; Dodd 1932). It was conceptualized as “the obligations of businessmen” (Bowen 1953, p. 6) “whose decisions and actions [are] taken for reasons at least partially beyond the firm’s direct economic or technical interest” (Davis 1960, p. 70). The concept evolved over time to encompass more specific responsibilities including economic, legal, ethical, and discretionary (Carroll 1979; Hill et al. 2007). Nowadays, the study of CSR has expanded significantly (Banerjee 2007; Garriga and Mele 2004; Margolis and Walsh 2003; Palazzo and Scherer 2006), offering today a wide array of approaches (Aguinis 2011; Carroll 1999; McWilliams and Siegel 2001; Waddock 2004; Waldman et al. 2006), each seeking to unveil particular mechanisms (Pasricha et al. 2017). Nonetheless, a universal definition of the meaning of CSR remains unreachable (McWilliams et al. 2006; Godfrey and Hatch 2007) primarily due to the wide-spreading variety of issues it covers (Okoye 2009) and the lack of clear empirical modeling (Godfrey and Hatch 2007), leaving this important concept as fuzzy and contested (Amaeshi and Adi 2007). Moreover, a study conducted by Okoye (2009) tested the concept of corporate social responsibility against the Essentially Contested Concepts (ECC) theory proposed by Gallie (1956) and confirmed that the term is in fact a concept that leads to enduring inconclusive debate.

Moreover, behind any corporate social responsibility initiative is an individual, whose personal values and motives are instrumental to establish social responsibility processes and initiatives (Hemingway and Maclagan 2004). From the perspective of the individual, the concept of social responsibility entails being ethical or acting responsibly (Victor and Cullen 1988), in a socially conscious way (Votaw 1972). Indeed, personal values, ethics, and social responsibility are intertwined (Joyner et al. 2002), and individual ethical principles of ‘doing what is right’ for society (Garriga and Mele 2004) can make a difference (Kahle et al. 1988). As such, individual lenders engage in lending activities that allow them to combine “financial objectives with their social values” (Munoz-Torres et al. 2004, p. 200) or “mix money with morality” (Diltz 1995, p. 64). Western societies, in particular, place higher emphasis on individuality and autonomy (Hofstede 1980), which highlights expectations of personal responsibility (Hill et al. 2007). By the beginning of the century, socially responsible investment had grown to reach one in every 8 dollars in financial markets (Laufer 2003). Furthermore, between 1995 and 2009, socially responsible investment products increased fivefold, from $639 billion to 3 trillion dollars (Social Investment Forum 2009), further highlighting its importance.

However, individuals driven by higher ethical purposes will not act altruistically in all situations. A recent study by Vitell (2015) demonstrates that individuals might alter their behavior towards ethical practices that lean more on individual rather than societal good, when the effect of an action provides an immediate positive effect on herself or himself, and the effect on society seems more distant or uncertain. Nonetheless, research suggests that even in trying conditions, the objective of well-intentioned ethical individuals remains to simultaneously achieve a financial return, while satisfying social responsibility goals (Hill et al. 2007) or an altruistic motive (André et al. 2017; Hemingway and Maclagan 2004). The question remains, however, to what extent strategic or altruistic factors influence motives by prosocial lenders.

Ethical Lending in Prosocial Crowdfunding: A Two-Pronged Approach

Evidence indicates that lending decisions are made both with rational processes (e.g., Simon 1959) and by relying on emotions and intuition (e.g., Bazerman and Moore 2012). However, the available literature has mostly examined more traditional forms of financing (e.g., institutional investing, stock purchase, venture capital, angels, etc.), leaving a void in our understanding of the dynamics and mechanisms that drive online lending-based crowdfunding. Nonetheless, crowdfunding may be the most intriguing setting yet as its participants are largely amateurs in business or entrepreneurial finance—a striking difference from most other lender segments. In this study, we add to our knowledge of the crowdfunding phenomenon by seeking to understand what drives lenders’ decisions in the specific context of ethically motivated, prosocial crowdfunding.

We draw upon social responsibility frameworks available in the literature which highlight lenders’ decision-making between financial and emotional return expectations (Nielsen and Riddle 2010). Specifically, in this study, we follow the conceptualization of duality presented by Hemingway and Maclagan (2004) to examine to what extent crowdfunding lenders on socially motivated crowdfunding platforms base their entrepreneurial lending decisions on altruistic or strategic motives (see Fig. 2).

Fig. 2
figure 2

Reproduced with permission from Hemingway and Maclagan (2004)

Social responsibility framework.

Full size image

Guided by this theoretical lens, we first examine lending-based prosocial crowdfunding from a strategic financial motive. Following this perspective, we expect that a crowdfunding lender, similar to other early-stage lenders, would examine ventures seeking funding based on signals of quality that guarantee success (Fiet 1995; Fried and Hisrich 1994; Mollick 2014; Tyebjee and Bruno 1984). In this process, lenders would seek to determine loan quality before committing. Research suggests that entrepreneurial financiers look for signals indicating the quality (e.g., sales, default or risk rates) of a loan in order to minimize losses and maximize returns (Nagy and Obenberger 1994). Therefore, we hypothesize that lenders use these signals to assess perceived risk and expected return, which play a critical role in whether or not a financier decides to fund an entrepreneur (Ganzach 2000; Lange et al. 2003; Tyebjee and Bruno 1984; Van Osnabrugge 2000).

Hypothesis 1

There will be a positive relationship between signals of loan quality and entrepreneurial success in lending-based prosocial crowdfunding.

However, lenders may also act in an ethical manner that adheres to higher order principles which are more altruistic in nature. This perspective aligns with literature in entrepreneurial financing, which suggests it is not uncommon for early venture financiers to offer funding driven by subjective criteria (Aernoudt 1999; Allison et al. 2013, 2015; Baty and Sommer 2002). Early-stage financiers often may want to assist certain people (Cardon et al. 2009; Mitteness et al. 2012; Sudek 2006) or support a specific product/service/sector (Brettel 2002, 2003; Harrison and Mason 2007). Lending in such a manner demonstrates a commitment to acting ethically, altruistically, and idealistically. Therefore, we hypothesize that lenders would provide crowdfunded loans based on a personal connection with an entrepreneur and/or venture, taking precedence over or in combination with the potential economic value of the deal—particularly in contexts which facilitate this behavior such as lending-based prosocial crowdfunding platforms. By taking this perspective, we expect lenders would rely greatly on altruistic criteria including helping those in need.

Hypothesis 2

There will be a positive relationship between the entrepreneur’s demonstration of need and entrepreneurial success in lending-based prosocial crowdfunding.

Method

Sample

This study uses data collected directly from Kiva’s website. Kiva is a non-profit organization that was founded in 2005 with a global mission to “connect people through lending to alleviate poverty” (Kiva 2017). We chose to use Kiva as it gave us an ideal platform to test the effects of both financial and altruistic motive. In other words, it is not fully altruistic as a donation-based platform (e.g., Gofundme), but it is also not a purely financially motived donation-based platform (e.g., Propser). Instead, Kiva occupies a middle ground which allows us to test financial and altruistic motives simultaneously.

Borrowing entrepreneurs and crowdfunding lenders from all over the world can sign up through Kiva’s online platform to either borrow or lend funds. An example of a borrowing entrepreneur’s page can be seen in Online Appendix A. Each entrepreneur goes through a field partner, who works with Kiva, to secure financing. Operating through the field partner, entrepreneurs can be backed by lenders from anywhere in the world, who voluntarily sign up with Kiva to lend funds to whomever they want. Different from other forms of crowdfunding (i.e., other lending platforms), lenders on Kiva receive no interest on the loan, only receiving the principal back at best, while risking total loss if the borrowing entrepreneur defaults.

Entrepreneurs on Kiva agree upon terms to repay the loan through the field partner. The repayments are collected by the field partner and the funds are then funneled back to Kiva lenders. Every step of this process is completed online through the Kiva platform. This concept of lending with no financial gain possible, only risk of loss, is one distinguishing characteristic of a handful of online crowdfunding platforms, most notably Kiva. This structure makes the Kiva platform an ideal setting to test our hypotheses of whether crowdfunding lenders use strategic or altruistic motives in a complementary manner, or whether the two perspectives are actually at odds with one another.

The sample used in this study includes public data available on Kiva’s website between 2008 and early 2013. A scraper program was created to collect these data from what is publicly available on the Kiva website. The scraped data include data on borrowing entrepreneurs, their projects, field partner intermediaries, and the countries in which this occurs. After dropping observations with incomplete data, our final sample includes 146,218 project loans which we use to test our hypotheses.

Measures

We use two methods to assess funding success. First, we use logistic regression for our analyses, where our dependent variable is a dichotomous variable indicating whether the entrepreneur’s loan got fully funded. Parameters estimated from the logistic data model indicate the direction of the effect of each explanatory variable on the response probability of the loan requested being fully funded. Second, we use linear regression to assess the relationships when our dependent variable is amount funded. This allows us to determine the effects of our variables of interest on the dollar amount that was actually funded to the entrepreneur.

Dependent Variable

We measure funding success in two different ways. First, fully funded is operationalized as a dichotomous variable with 1 indicating that the entrepreneur had their venture fully funded, and 0 indicating that the entrepreneur seeking venture funding did not receive the full amount requested. Our dependent variable was identified because of the definitive importance that obtaining start-up capital has on new ventures, as identified in the literature (Hellman and Puri 2002; see also Hall and Lerner (2010) for a summary of the literature). Second, we measure funding success using amount funded. Often this is measured as a ratio of asked amount versus received amount, but since only loans which are 100% backed receive funding (funds are refunded to the lender if full funding for a loan request is not achieved), this approach was not possible. Therefore, we measured dollar amount as the total dollars raised through Kiva to back the field partner loan (i.e., 0 if they failed to get full funding or the dollar amount they asked and received through Kiva). Total amount raised ranges from $0 to several thousand dollars. This variable was transformed using the natural logarithm to account for skewness.

Independent Variables

Independent variables which provide information on financial appeal (strategic) or social appeal (altruistic) were selected from the pool of readily available information on Kiva. This seems intuitive as this is the same pool of information which lenders use to assess and choose which projects to fund. Using the informational signals available, we chose two measurements of financial appeal which would appease a strategic motive. First, at the field partner level, we examine risk rating which indicates the risk rating assigned to the field partner overseeing the loan. Field partner risk rating ranges from 0 to 5 stars in increments of 0.5 (0, 0.5…,4.5, 5.0) which our scraper converted to 0 to 10 (0, 1…,9, 10), with 10 indicating the lowest risk (i.e., better financial option). Second, at the project level, we examine whether the individual loan had loss protection using a dichotomous variable, with 1 representing a loan which had protection from loss due to default or currency exchange issues, and 0 if the loan did not have such protection.

Similarly, we chose two measurements of social appeal which would entice those with more altruistic motives. First, at the field partner level, we examine the role of social performance badges on influencing the emotional decision of lenders to fully fund the entrepreneur. These badges are graphic icons which appear on each entrepreneur’s profile with the field partner that manages the loan. These badges indicate the focus areas of the field partner which handles the loan and the type of entrepreneurs/projects it supports. We coded this as a dichotomous variable that indicates 1 if the field partner holds a badge, which denotes focused support for those in need and zero otherwise. Second, at the project level, we examine the presence of altruistic-appealing keywords in the descriptive narrative that accompanies the entrepreneur’s profile. The data regarding the descriptive narrative that accompanies the entrepreneur’s profile were coded, resulting in an altruistic narrative variable that indicates the number of times a keyword is mentioned in the entrepreneur’s profile narrative. This technique has been previously used by other scholarly work in crowdfunding (e.g., Allison et al. 2015; Parhankangas and Renko 2017). Our keywords were motivated by previous studies examining prosocial language (Frimer et al. 2015; Pietraszkiewicz et al. 2017) and include words such as unemployed, disabled, vulnerable, and poverty, to name a few. A full list of the terms used in this study which may motivate altruistic funding actions can be seen in Online Appendix B. As a robustness check, we also created several different combinations and subgroups of these words ranging from 5 to 75 terms. Using these various groupings did not change our results.

Measurement of these constructs, particularly at the field partner level, is inherently messy and concerns about their validity have been noted, therefore we offer additional information about the field partners in this study in Table 1. The first column of Table 1 represents some variables of our 81 field partners with means and averages (columns 2 and 3) for all. The last few rows show the top industries funded and countries funded across field partners. We parsed out high and low financial appeal (by risk rating at the field partner level) in the middle section. As displayed, highly rated field partners have lower delinquency rates, refund rates, and are more often connected to a network of affiliates. These differences are significant at the 1% level and offer a deeper view into the field partners, the projects they are intermediaries for, and offer validation for our financial appeal construct.

Table 1 Field partner statistics
Full size table

Similarly, we parsed out high from low social appeal field partners based on whether they possessed an appropriate badge indicating that they serve more vulnerable populations. Indeed, field partners which possess a badge serve in countries with higher infant mortality rates, lower life expectancies, and lower literacy rates, which would indicate that they are indeed catering to more vulnerable populations. These differences are also significant at the 1% level. This supports our supposition that a badge is indeed an appropriate proxy for helping those in need. The significant difference between motives also shows that field partners, and subsequently lenders, cater to different groups (demonstrating a trade-off between the two motives). This table also offers preliminary support for our hypotheses as financially appealing field partners garner more loan dollars for the entrepreneurs they back, and possessing a badge has a significantly lower funding success rate. Correlation analyses support these findings. We also conducted further robustness checks of the financial and social motives at the project and field partner level. These are discussed in our “Results” section.

Control Variables

We control for several types of variables including entrepreneur, venture, and field partner characteristics. We use these control variables to account for other information available to the lender from the crowdfunding platform.

First, we control for the amount requested. This is a continuous dollar amount that is transformed using the natural logarithm to curtail skewness. We also control for whether the entrepreneur is in a group with a dichotomous variable, where 1 represents that they are part of a team of entrepreneurs and 0 indicates they are not. Next, we control for both the length of the anticipated repayment term (in months) and whether the repayments were to take place at irregular repayment intervals, or if they were regularly scheduled repayments. The repayment schedule is coded as a dichotomous variable with 1 representing an irregular repayment schedule and 0 indicating a regular repayment schedule for the entrepreneur to pay back the funds lent. We also control for various field partner characteristics, including the number of entrepreneurs funded by the field partner previously, the total number of dollars lent by the field partner previously, and the average loan size in dollars and average loan term in months. Prior literature in entrepreneurship and strategy on field partner characteristics in crowdfunding research is very scarce, although a recent study by Allison et al. (2013, 2015) follows a logic similar to ours. Furthermore, we control for the industry, year, and country effects.

Results

Table 2 provides the descriptive statistics for all variables used in this study. Descriptive statistics include the means, standard deviations, and bivariate correlation coefficients for all variables in our study derived from our sample of 146,218 loan observations. There is a high rate of fully funded ventures (approximately 99%), however, this is not surprising as Kiva prides itself on the high funding and low default level of its participants.

Table 2 Descriptive statistics and correlations
Full size table

Results from our logistic regression analyses are reported in Table 3. The analyses involved testing of a baseline model, followed by individual tests of our hypotheses, and finally a full model with all predictor variables (Model 6).

Table 3 Logit regression results: fully funded
Full size table

The first theoretical framework used in this study focused on strategic criteria for providing funding for entrepreneurial ventures. Hypothesis 1 proposed a positive relationship between signals of loan quality and entrepreneurial funding success in lending-based prosocial crowdfunding. Our study focused on two signals for the analyses. As seen in Model 2 of Table 3, the coefficient on risk rating is significant in a positive direction (β = 0.39, p < 0.01). This result indicates that a field partner with a better risk rating increases the chances the entrepreneur has of being fully funded by lenders. Regression results for the hypothesized relationship between protection from entrepreneur loss coverage and being fully funded, as displayed in Model 3 of Table 3, are also positive and significant (β = 1.84, p < 0.01). Both results consistently support Hypothesis 1 at both the field partner and project level.

The second hypothesis predicted that entrepreneurial lenders would invest altruistically responding to and investing in entrepreneurs who demonstrate need. Empirically, we operationalized this hypothesis using two variables. Our first variable indicates whether the entrepreneur’s profile denotes a field partner with a socially appealing badge, which we hypothesized would offer the entrepreneur a higher likelihood of being fully funded. As seen in Model 4 of Table 3, results from our analyses lend support for the opposite of what we had hypothesized (β = − 0.69, p < 0.01) suggesting that such a badge on the entrepreneur’s profile may deter financial lenders to fully fund the project. Our second variable indicates an altruistic entrepreneurial narrative, denoting need and therefore seeking to appeal to the altruistic nature of lenders. As seen in Model 5 of Table 3, our results indicate entrepreneurs who communicated need via their narrative were actually less likely to get their loan fully funded (β = − 0.05, p < 0.01). Both, the signal from the field partner and the project consistently reject Hypothesis 2, suggesting that altruistic factors have a contrary effect to what we predicted. All results hold in Model 6, the full model. As a robustness check and to quell concerns with the disproportionality of our 1/0 dependent variable, we ran Firth logit models (Firth 1993). Firth models account for such disproportionate distributions in a binary dependent variable (Firth 1993; King and Zeng 2001). The results remained unchanged with the Firth models.

In addition to examining whether projects were funded or not, we also examined the dollar amount funded which captures not only whether the projects were funded, but also how much was funded. Unlike other crowdfunding platforms, the amount of funding dispersed to borrowing entrepreneurs on Kiva cannot exceed the amount requested. Furthermore, if a loan does not get full backing, the lenders are refunded their money and the loan request is eventually removed. Therefore, some projects ended up with no funding through Kiva while the remainder (99%) achieved full funding of the amount requested. In our sample, this amount varied from $0 to $20,700. In Table 4, we used linear regression to set up a baseline model (Model 1) followed by testing each hypothesis individually, before examining all the variables together.

Table 4 Regression results: amount funded
Full size table

Models 2 and 3 of Table 4 test the strategic motives by examining signals of financial appeal. In Model 2, the risk rating of the field partner is incorporated. As expected, the coefficient is positive and significant (β = 0.02, p < 0.01) suggesting that if borrowers go through high-rated field partners they can achieve higher amounts of funding. Model 3 incorporates whether the loan is covered for loss. Once again, the coefficient is positive and significant (β = 0.06, p < 0.01) suggesting that individual projects which offer financial protection to the lender can achieve higher funding amounts. These results are consistent with our hypotheses as well as with our logistic regression results.

Models 4 and 5 test our altruistic motive hypotheses. Model 4 examines whether a borrower going through a field partner with a social badge received higher funding amounts. The coefficient is negative and significant (β = − 0.03, p < 0.01) suggesting that going through a field partner with such a badge is a detriment to funding. Model 5 examines the narrative of the entrepreneurial borrower. The coefficient is again negative and significant (β = − 0.01, p < 0.01) suggesting that a more need-driven narrative is actually a detriment to the amount of funding received. These findings are consistent in the full model (Model 6).

Additional Analyses

To further illustrate our findings, we compiled a figure with projects categorized by either high or low financial appeal (risk rating and loss coverage), along with high or low social appeal (badge and narrative). As depicted in Fig. 3 and consistent with our regression results, projects with high financial appeal outperform those with high social appeal. Specifically, projects with high financial appeal and low social appeal have the highest funding success rate (99.54%) and a higher dollar amount funded ($747.87), whereas projects with high social appeal and low financial appeal have the lowest success rate (96.81%) and a lower dollar amount funded ($683.01). ANOVA analyses and supplemental t-tests confirm significant differences between groups across both the signals offered (strategic and altruistic) and funding outcomes (success and dollar amounts). Notably, though, we also find that projects high on both financial and social appeal (Group 1 in Fig. 3) received the highest average amount of funding. This result provides some nuance to our findings about the prevalence of strategic lending motives, while being consistent with research suggesting that the overarching objective for ethical individuals is to simultaneously achieve a financial return, while satisfying social responsibility goals (Hill et al. 2007) or an altruistic impulse (André et al. 2017; Hemingway and Maclagan 2004).

Fig. 3
figure 3

Success of loan groupings—main effects

Full size image

To further parse out financial versus social appeal, we assessed the language of each proposal in terms of ability to pay—another metric of financial soundness of the loan (see Fig. 4). We plotted out high versus low ability to pay, along with high versus low social appeal. Similar to our results above, higher social appeal is associated with lower funding rates, but higher funding amounts. Once again, ANOVA analyses and supplemental t tests confirm significant differences between these groups across both of the signals offered (strategic and altruistic) and funding outcomes (success and dollar amounts). Language indicating the ability to pay is not significantly correlated with loss coverage (0.01) and risk rating (0.07). Furthermore, re-running regression analyses with ability to pay in the equation indicates that ability to pay is positively related with both funding amount (β = 0.03, p < 0.01) and funding success (β = 0.42, p < 0.01), while previous results remain unchanged.

Fig. 4
figure 4

Success of loan groupings—robustness check

Full size image

Additionally, we deconstructed motives at the project level [loss coverage (financial) and narrative (social)] versus field partner level [risk rating (financial) and badge (social)] and assessed the differences between financial and social motives. Indeed, loans that have loss protection (greater financial appeal) have less need implying narratives (less social appeal) (Mean = 3.47) while loans without loss protection (less financial appeal) have more need implying narratives (greater social appeal) (Mean = 3.69) with the mean being significantly different at the 1% level. Similarly, loans through partners with badges indicating they help vulnerable populations (greater social appeal) have less favorable risk ratings (less financial appeal) (Mean = 5.43) while loans through field partners without such a badge (less social appeal) have more positive risk ratings (greater financial appeal) (Mean = 6.53) with the mean difference again being significant at the 1% level. Parsing loans in this way offers additional support for both our measures and our findings.

Overall, our findings suggest that strategic motives have a positive effect on the loan being fully funded, while altruistic motives have a negative effect. While it is not surprising that lenders invest strategically following signals of quality, it is surprising that, given the prosocial context in nature of the Kiva platform, altruistic motives appear to be detrimental. Nonetheless, although our results seem to contrast the results of Allison et al. (2015) who found that lenders on Kiva act pro-socially, we also find that projects high on both financial and social appeal receive the highest average amount of funding. This composite result is an important contribution to the literature.

Discussion

Securing funding is one substantial obstacle many entrepreneurs face (Ebbers and Wijnberg 2012; Mata 1994). While traditional financing sources such as family, bank loans, and venture capital remain common, many small entrepreneurs are increasingly turning to online crowdfunding for start-up capital (Polzin et al. 2018). Connectivity has made this financing option relatively easy, expanding the reach to lenders from all over the world, as well as increasing loan quantity through crowdfunded lending. Although the use of such platforms has spiked, our knowledge remains scant regarding lenders’ motives on crowdfunding platforms (Bretschneider and Leimeister 2017). We contribute to the limited literature on crowdfunding research as it pertains to business ethics implications by examining both strategic and altruistic motives in the context of lending-based prosocial crowdfunding. Specifically, this study makes a contribution at the intersection of business ethics and social responsibility. Indeed, it highlights the critical role of social responsibility at the individual level to allow lenders to combine moral values with financial objectives, providing some insight as to whether the primary motive to lend in a prosocial context is based on strategic or altruistic indicators, and the extent of that drive (i.e., intensity).

Our results indicate that crowdfunding decisions on Kiva tend to be primarily influenced by strategic, rather than altruistic motives. While there is evidence that individuals do not always rely on strategic processing, but rather fall victim to cognitive hurdles (Bazerman 1990), we find support for a certain rationality regarding lending decisions in prosocial crowdfunding. Notably, our results suggest the opposite direction for our hypothesis proposing that lenders make altruistic lending decisions in a prosocial context. One explanation for this may be that while lenders on Kiva are pro-socially motivated, they are also savvy enough to realize that a low-quality entrepreneur may actually deter them in their mission to help others by potentially defaulting on repaying the loan. In other words, by choosing high-quality entrepreneurs, lenders can ensure that their funds are paid back. Lenders can then reinvest their money on other high-quality entrepreneurs, further extending Kiva’s prosocial mission, while also satisficing their individual social responsibility goals (Hill et al. 2007). This idea is also consistent with André et al. (2017) who found that successful rewards-based crowdfunding, another type of crowdfunding, relies on reciprocity mechanisms. Future research could thus examine whether and to what extent lenders continuously reinvest their money when lending on prosocial crowdfunding platforms, such as Kiva.

Our findings are particularly interesting as they somewhat contrast those of Allison et al. (2015) who found, using a much smaller sample of only fully funded loans from Kiva, that altruistic narratives decreased the time to funding while strategic language increased the time to funding. Combining their results with ours would suggest that while altruistic narratives may increase the speed at which full funding is achieved (for ventures that are funded) they do not necessarily lead to full funding to start with (when examining the universe of entrepreneurs seeking funding). Furthermore, while strategic language may lead to increased time to full funding, strategic quality signals lead to a higher full funding rate.

Our study suggests that a social responsibility framework contributes to a deeper understanding of the dynamics of lending-based prosocial crowdfunding. Indeed, our explanation that crowdfunders tend to lend to high-quality entrepreneurs, which may lead them to continually reinvest their money in such crowdfunding endeavors, is consistent with social responsibility and altruistic ethical motives based on reciprocity (André et al. 2017). Furthermore, our finding that projects high in financial and social appeal receive the highest amounts of funding is also consistent with a social responsibility lens (Hemingway and Maclagan 2004). Interestingly, when breaking our sample down by type of projects being funded, we find that the projects that receive the highest average amount of funding are projects that are high in both social and financial appeal. Such a finding is noteworthy as it suggests that if entrepreneurs are both high quality and socially appealing, they are likely to receive higher amounts of funding.

Taken together, our results indicate a complex relationship between financial and altruistic motives. Both constructs play a unique role in prosocial lending in crowdfunding. Specifically, we find that financial metrics may be particularly important for securing financing (i.e., strategic motives lead to full funding)—while altruistic signals are the main driver for “intensity” of the decision and motive average funding amount—albeit at lower success levels. These results highlight the nuanced, but critical role that individual ethics play in business decisions in a prosocial lending-based crowdfunding context. Whereas metrics determine the overall success of a venture on Kiva, altruistic signals on the entrepreneur’s profile seem to contribute to intensity of the lender’s ethical drive to support higher loan amounts for those entrepreneurs whose profile denotes the most social appeal.

Finally, prior research indicates that individuals driven by ethical purposes may alter their altruistic behavior when the effects on society seem distant or uncertain (Vitell 2015). Our results show that lenders in a prosocial lending context actually combine moral and social values with financial objectives, consistent with research on ethical individuals in other contexts (Hill et al. 2007). Specifically, our findings suggest that strategic motives drive the successful funding of the loan, but when considered in combination with the entrepreneur’s ‘social appeal’ in the eyes of the individual lender, the entrepreneur is also likely to receive higher amount of funding, on average. These results are noteworthy because by studying whether the primary motive to lend is based on strategic or altruistic indicators, and the extent of that drive (i.e., intensity), this study further advances our understanding of the concept of social responsibility—acting responsibly or being ethical—from the perspective of the individual. Although social responsibility was initially considered at the intersection of social-service and profit-making (Berle 1931; Dodd 1932), it is increasingly becoming an influential investment driver in current financial markets (SIF 2009; Laufer 2003). Moreover, as Hemingway and Maclagan (2004) suggest, behind any corporate social responsibility initiative lies an individual, who possesses personal values and motives that drive their decision-making. Given our limited understanding about the motives that influence lenders to fund entrepreneurial endeavors (André et al. 2017; Bretschneider and Leimeister 2017; Cholakova and Clarysse 2015), especially in newly emerging financial platforms, this study further highlights the role of personal responsibility (Hill et al. 2007) in business ethics through responsible investment in a prosocial context.

Practical Implications

Our results have practical implications for small entrepreneurs and lenders. For small entrepreneurs around the world, being aware of the factors that influence lending motives could be an important competitive advantage. Awareness of primary lending motives may be pivotal not only to secure funding, but also to identify both lenders and intermediaries that may be more or less appropriate to seek out to help further the entrepreneur’s business. Furthermore, lenders need to be aware of the factors and motives that are swaying their lending decisions. In other words, is the pool of lenders on a certain crowdfunding platform following signals of quality and risk reduction? Or, do they throw caution to the wind and make their decisions based on more altruistic factors regardless of the monetary payback? The answers to these questions would enable small entrepreneurs and lenders to be more efficiently matched—benefiting both parties.

Limitations and Directions for Future Research

Caution should be used before generalizing these findings to other crowdfunding contexts. Although Kiva is one of the largest crowdfunding platforms, and the largest lending-based prosocial platform to date, lenders on Kiva may represent a unique segment of the crowdfunding universe. As such, a fruitful area of future research would be to further examine differences across types of crowdfunding platforms (i.e., equity-based, reward-based, and donation-based), comparing and contrasting what factors drive lenders and/or entrepreneurs on each of these platforms. For example, do platforms inevitably cater to lenders who lend either strategically or altruistically, or instead do platforms comprise a continuum with a range of motives? Using a social responsibility framework, these and other questions could be answered at the organizational or individual level. The available information about crowdfunding platforms to date remains limited and differs due to key differences between and among platforms, making comparisons between platforms difficult. This lack of knowledge on the dynamics of crowdfunding platforms exacerbates the relevance of this present study and calls for more research on the topic to more effectively inform entrepreneurs seeking financing, crowdfunders seeking to lend, and policy makers interested in motivating both lenders and entrepreneurs.

Another limitation lies in the data used in this study, as in most crowdfunding research to date. The data for this study were gathered from what is publicly available on Kiva’s website. Although this is the same information that is readily available to lenders, these data may present a limited window into the mechanisms driving the motives behind lending-based prosocial crowdfunding. Continued research is necessary to capture more detailed data from both the funders and entrepreneurs involved in crowdfunding to understand its true complexity. For example, with regard to the handful of loans that failed, why did they fail to receive full funding? Further exploration of loans that were successful, but that defaulted is critical to accrue a more complete understanding. Relatedly, future research could also differentiate liquidity from strategic defaults (Giroud et al. 2012).

Finally, surveys and interviews could also be utilized to capture more elaborate perspectives from both sides (for example, see Cholakova and Clarysse 2015; Polzin et al. 2018). Similarly, longitudinal data examining various tendencies and subsequent success would also significantly broaden our understanding of crowdfunding lending decisions. In addition, while we examined social responsibility from the perspective of the lenders, further studies could also follow the work of Azmat and Samaratunge (2009) who explored socially responsible entrepreneurs, and Aribi and Arun (2015) who explored financial institutions. Specifically, how are entrepreneurial ventures or financial intermediaries who exemplify social responsibility and ethical actions perceived by lenders and other stakeholders?

Conclusion

Crowdfunding is an increasingly prevalent and unique channel of entrepreneurial finance. With the growing speed and reach of the internet, crowdfunding is now “only one click away.” By analyzing the content of a major lending-based prosocial crowdfunding platform, Kiva, we were able to examine a unique segment of lenders, delve into the subtleties of this type of platform, and expand our knowledge on linkages of business ethics with individual social responsibility. Specifically, we answered calls to extend theory as it pertains to crowdfunding (McKenny et al. 2017), in the context of the strategic and altruistic ethical motives that drive lenders to engage in crowdfunding lending (Bretschneider and Leimeister 2017). We found that lenders in prosocial crowdfunding tend to follow strategic over altruistic motives, while ventures high in both social and financial appeal tend to attract the highest amounts of funding. Whereas strategic motives determine funding success, altruistic motives may drive the intensity of the decision via average funding amount.