**Human Capital, Social Capital, and Farmers' Credit Availability in China: Based on the Analysis of the Ordered Probit and PSM Models**

#### **Jiaojiao Liu <sup>1</sup> , Gangren Zhang 1,2, Jun Zhang <sup>3</sup> and Chongguang Li 1,\***


Received: 3 January 2020; Accepted: 17 February 2020; Published: 20 February 2020

**Abstract:** Rural credit is very important to the increase of farmers' income and the development of rural economy, and it has attracted wide attention from scholars. Many scholars have paid attention to the impact of social capital on farmers' credit availability, but the research conclusions have not yet been unified. In addition, human capital is also one of the important factors that scholars pay attention to. However, the research mainly focuses on farmer education and pays less attention to their health. Based on the China Household Income Project (CHIP2013) database, we evaluated the impact of human capital (education and health of farmers) and social capital on the credit availability of farmers. To ensure the robustness of our results, we used both the ordered probit model and the propensity score matching (PSM) model to carry out the estimations. Therefore, the study not only improves the research framework of the impact of human capital on farmers' credit availability, but also uses a more accurate method to estimate the net impact of social capital on farmers' credit availability. The results showed that, firstly, in terms of human capital, farmers' educational and health levels have a significant positive impact on their formal credit availability, but no significant impact on their informal credit availability. In particular, farmers with a high school education or above are more likely to obtain a formal loan. Secondly, in terms of social capital, interpersonal relationship capital and political relationship capital are beneficial for farmers obtaining loans from formal and informal channels. Organizational relationship capital only has a more significant positive impact on the informal credit availability of farmers. These results imply that formal financial institutions not only pay attention to farmers' human capital but also their social capital to reduce the risk of lending. However, informal lenders, that is, relatives or friends, pay more attention to the social capital of farmers.

**Keywords:** human capital; social capital; credit availability; propensity score matching; China

#### **1. Introduction**

The development of the rural economy cannot do without the support of rural finance [1]. A relatively perfect rural financial market can significantly improve farmers' technical efficiency and increase their income and consumption [2,3]. However, at present, the credit constraints on farmers are still relatively serious in China, especially the credit constraints from formal financial institutions. Further research also found that the welfare of farmers with credit constraints has been significantly reduced. Kumar et al. [4] found that credit constraints had a negative impact on farmers' health and education expenditures, food consumption, and agricultural investment. Therefore, researchers have

focused on methods to reduce the credit constraints on farmers or improve their credit availability. Among them, the impact of material capital and social capital on the credit availability of farmers has received considerable attention.

Relevant research shows that there are many reasons why farmers are subject to credit constraints, such as the lack of effective mortgages and a perfect rural credit system and the asymmetric information between borrowers and lenders [5]. Material capital has a positive impact on farmers' credit availability because it can be used as an effective mortgage to restrain the potential default behavior of farmers. Furthermore, due to "limited liability", banks often decide whether to lend to farmers and the specific loan amount based on the farmers' wealth owned. Alternatively, they may set a higher credit threshold (e.g., mortgage and guarantee) to ensure that the borrower has a certain ability to repay to reduce their lending risk [6]. Consequently, wealthier families often have more access to bank loans because they have the necessary mortgage for loans [7]. Moreover, farmers who have a higher income always have a higher credit rating, and a bank will treat them as quality customers and lend them more money [8]. Furthermore, Xu and Yuan [9] argued that wealthier farmers have more additional capital to invest in their social networks and expand financing channels. Their study found that, compared with the least wealthy 10% of farmers, the credit availability of the wealthiest 10% of farmers is significantly better.

However, the income of farmers is generally low in China, and the material capital that can be used as an effective mortgage is generally insufficient. Social capital can be used as a substitute or supplement to material capital and can reduce the negative impact of insufficient material capital on farmers' credit availability [10]. Therefore, the impact of social capital on farmers' credit availability is also one of the important factors to which many scholars pay attention. However, studies on farmers' credit availability mostly used a binary variable ("whether to obtain loans?") or specific financing amounts [11,12]. However, with the development of the economy and the improvement of farmers' income and consumption, the borrowing amount of farmers has increased significantly, while the credit constraints felt by farmers have not eased and even increased [8]. This may be because the credit demand of farmers is also increasing. Consequently, it may not be appropriate to use a binary variable (yes or no) or a specific borrowing amount to express farmers' credit availability. The dependent variable selected in this work also considered the credit demands of farmers, which may better express the credit availability of farmers.

In terms of human capital, most studies about the impact of human capital on the credit availability of farmers have focused on their education and less on their health [13,14]. However, human health is also one of the most important components of human capital [15]. Therefore, in this study, we estimated the impact of both farmers' education and health on their credit availability and improved the research framework of the impact of human capital on farmers' credit availability.

Furthermore, there is a lack of consistency among the research conclusions regarding the impact of human and social capitals on farmers' credit availability. To ensure the robustness of our results, we used both the ordered probit model and the propensity score matching (PSM) method to carry out the estimations. For some important human and social capital variables (e.g., participation in cooperative organizations) that are self-selective, we re-estimated their impact with the PSM method to avoid the impact of sample selection bias on the estimation results.

The remainder of this paper is arranged as follows. The second part mainly introduces the credit situation of farmers and the cultural background of China. The third part mainly reviews the literature on human and social capitals and the credit availability of farmers. The fourth part explains the ordered probit and propensity score matching models to analyze the net effect of human and social capitals on the credit availability of farmers. It also explains the data source and the data type used for the estimation. The fifth part empirically analyzes the impact of human and social capitals on the credit availability of farmers. The last part summarizes and discusses the main findings and draws some suggestions to improve the credit availability of farmers.

#### **2. Research Context**

China is a large agricultural country. However, China's agricultural production is small-scale and decentralized, with a low production efficiency and rising production costs. Moreover, agricultural production also faces both natural and market risks. Therefore, farmers' agricultural income is low and unstable. In such circumstances, financial institutions are generally unwilling to lend to farmers.

Furthermore, the construction of the rural financial system is not perfect in China. There are few financial institutions and an uneven distribution in rural areas. Financial services are more traditional and single, which cannot effectively meet the needs of farmers. Besides, rural finance also lacks a risk-sharing mechanism. Therefore, the phenomenon of "de-ruralization" of financial institutions is serious [16], and farmers are severely constrained by credit.

However, at present, the credit constraints on farmers are still relatively serious in China, especially the credit constraints from formal financial institutions. Li et al. [17] made a survey of 1773 households in China, and the data shows that among the farmers with a borrowing demand, about 66.92% of the farmers are subject to credit constraints. He et al. [18] also conducted a field survey on the credit constraints of 1730 households in Shandong, Henan, and Guangxi provinces. They found that about 31.21% of the households were subject to credit constraints and could not obtain loans from formal and informal channels. Among the farmers who had received loans, only about 34.02% of them obtained loans from formal channels, and about 53.61% of them could only acquire loans from informal channels. Furthermore, compared with farmers who are not subject to formal credit constraints, the productive income of farmers with partial or complete credit constraints would decrease by 13.0% and 9.8%, respectively. Additionally, the non-basic consumption of farmers with partial or complete credit constraints would decrease by 14.8% and 12.5%, respectively [19].

China's rural residents are collective and closely related based on blood, kinship, and geography. Farmers live in the same place for a long time, forming certain social norms. All farmers will consciously abide by these unwritten norms. In addition, the spatial distance between families is close, and the communication between farmers is frequent. Therefore, the degree of information sharing between farmers is high, and the speed of information transmission is also fast. If someone violates these norms, they will be rejected by others and under the pressure of gossip. As an old Chinese saying goes, "good news never goes out, while bad news has wings". That is, the transmission speed of "bad news" is very fast, such as information that does not comply with the norms. Therefore, under social pressure, farmers may be more likely to comply with the norms. The unique rural culture in China implies the particularity of China's problems.

#### **3. Theoretical Analysis and Research Hypothesis**

Farmers' borrowing is a transaction between farmers, or between farmers and financial institutions. Transactions always have costs. Transaction costs include the cost of information search before the transaction, the cost of bargaining during the transaction, and the cost of supervision and execution after the transaction [20]. One of the reasons for the transaction cost is the information asymmetry between the borrowers and lenders. In the case of information asymmetry, borrowers may produce opportunistic behavior. To reduce or prevent the borrowers' opportunistic behavior, lenders need to spend more money to collect more information.

For human capital, farmers with higher education levels tend to have higher comprehensive qualities and a lower probability of opportunistic behavior. For social capital, the collateral function and information transmission function of social capital can reduce transaction costs such as the cost of information search between farmers or between farmers and banks. Therefore, both human capital and social capital can reduce the transaction costs between borrowers and lenders, and help farmers obtain loans.

#### *3.1. Human Capital and the Credit Availability of Farmers*

Human capital is formed by workers' investments, which reflect the knowledge, skills, and health level of workers [21]. It is an important factor in promoting economic growth and increasing the income of farmers [22]. Cheng et al. [23] found that human capital contributed 38.57% to the increase in farmers' income, among which the health and education of farmers played an important role. Studies have also shown that, due to "limited liability", lenders such as banks tend to lend more money to wealthier farmers [24]. Therefore, farmers with a higher human capital may be more likely to obtain loans. The measurement of human capital also differs in the literature. It is often measured by an individual's education, training, working seniority, health, and other indicators [25,26]. However, an individual's education and health are the two most important components of human capital [15]. Therefore, we chose farmers' education and health to express the human capital of farmers in this study.

Generally speaking, education can effectively distinguish high-ability from low-ability people [27]. Farmers with a higher level of education always have a higher comprehensive quality, and there are more or better employment and learning opportunities available to them to improve their income [22]. At the same time, Yi and Cai [28] claim that, compared with low-income farmers, high-income farmers tend to have a better repayment willingness and ability. This may reduce the potential default risk that banks and other lenders may bear. Finally, a higher level of education may make it easier for farmers to obtain loans.

As the saying goes in China, the "body is the capital of the revolution". A healthy body is an important carrier of other human capital components [29]. The health of farmers also has a significant impact on farmers' current income, even more so than the education [30]. Furthermore, farmers' health also indicates their future income and repayment ability. Farmers with a better health level can maintain and improve their income and repayment ability [31]. Therefore, the health of householders represents the credit risks of lenders to some extent; the healthier the householder is, the less risk the lender will face [10], and the healthier householders will have a higher probability of obtaining a loan. Yin et al. [31] found that the average health level of family members has a significant positive impact on the formal financing capacity of farmers. Based on this, we hypothesized the following:

**Hypothesis 1.** *Human capital has a significant impact on farmers' credit availability. The higher the education and health level, the higher the credit availability of farmers.*

#### *3.2. Social Capital and the Credit Availability of Farmers*

For social capital, there is currently no universally agreed-upon definition, but the definition described by Putnam is widely used. Putnam [32] points out that social capital is an organizational characteristic that can improve economic efficiency and people's income, which includes trust, norms, and networks. Heikkilä et al. [33] studied the relationship between individual social capital and credit availability in Uganda. They found that the importance of individual social capital to formal banks was significantly reduced because they valued a physical mortgage more greatly. However, social capital has a significant impact on the semi-formal and informal credit availability, especially for poor people, and for those in rural areas or areas with low general trust. These findings support the views of Liang et al. [6], who found that formal financial institutions have not yet taken farmers' social capital as the basis for lending. This results in social capital having no significant impact on the formal financing ability of farmers and only having a significant impact on their informal financing ability. However, van Bastelaer [34] argued that social capital could reduce the cost of incomplete information in financial transactions. Social connections between borrowers allow significant savings in terms of screening, mutual monitoring, and enforcement. This kind of interpersonal relationship is a central factor in ensuring repayment and is one of the important factors that lenders consider. Tan and Hu [24] also found that social capital could significantly improve the formal credit availability of farmers. When

social capital is increased by one unit, the probability of farmers being subject to credit constraints is reduced by about 20%.

There may be three main mechanisms for the impact of social capital on the credit availability of farmers. First, social capital owned by farmers can be used as a social mortgage, and its punishment and reputation mechanism can effectively restrict the behavior of farmers [35]. The countryside is a typical acquaintance society in China. Farmers live in a group for a long time and form some social norms. Once farmers violate the social norms, they come under social pressure from the group (e.g., relatives and friends), which causes a certain loss of their reputation and increases the cost of their default [12]. Moreover, the higher the social capital stock of farmers, the higher the cost of their default. This may give farmers a stronger incentive to repay on time to maintain or further enhance their social capital [36]. The high cost of default may also reduce the concerns of lenders such as banks, and then improve the credit availability of farmers.

Second, social capital also has the function of information transmission, which can reduce the information asymmetry between borrowers and lenders. With the low marketization degree of rural finance in China, the role of social capital becomes particularly important. Villages with higher trust levels and developed social networks have a higher information sharing level. In such a village, farmers' personal information is more fully disclosed [37], such as farmers' risk type, borrowing demand, repayment ability, and other information. To a certain extent, it may alleviate the adverse selection and moral hazard caused by information asymmetry between borrowers and lenders.

Third, social capital can also help farmers acquire more borrowing resources. Dinh et al. [38] argued that building strong ties with people of a higher social status could reduce credit constraints. Li et al. [17] found that one of the important reasons why farmers think they cannot obtain loans from banks was that they had no acquaintances at the banks. Farmers with relatives working in the financial sector tend to have more borrowing resources [39]. In addition, participation in credit cooperatives could significantly reduce farmers' credit constraints. Even for poor farmers, their borrowing opportunities also increase significantly [40]. Therefore, we hypothesized the following:

**Hypothesis 2.** *Social capital has a positive impact on farmers' credit availability.*

#### **4. Research Methodology**

This study mainly aims to answer the following question: "Do human and social capitals have a significant impact on the credit availability of farmers?" We mainly use an empirical analysis to test, including ordered probit and propensity score matching (PSM) models. However, it is worth noting that some variables representing human and social capitals are a kind of self-selection of farmers. These selections may not be random and may be influenced by the characteristics of the farmers themselves. Moreover, these characteristics may also affect the credit availability of farmers. If so, the general regression model cannot completely exclude the influence of other factors and obtain the net impact of these variables on farmers' credit availability. However, the propensity score matching (PSM) model can effectively control this nonrandom bias problem through a counterfactual estimation [41,42]. Therefore, we choose both the ordered probit and PSM models to estimate the impact of human and social capitals on the credit availability of farmers.

#### *4.1. Collection of Data*

The data used in this study came from the China Household Income Projects (CHIP2013) database, which was completed by the China Income Distribution Institute of Beijing Normal University and domestic and foreign experts in 2014. The CHIP project team took samples according to stratified random sampling and systematic sampling methods. They stratified the region according to the east, the middle, and the west, and then obtained samples according to the systematic sampling method. Finally, the sample covers the eastern, central, and western parts of China: 15 provinces; 126 cities; 234 counties and districts; and a total of 18,948 household samples, including 11,013 rural household samples, 7175 urban household samples, and 760 migrant household samples. Considering the research topic of this study, we mainly selected farmers who had applied for loans from 11,013 rural household samples. Then, we removed the samples with missing data and finally obtained a total of 3127 effective samples.

The database collects the basic characteristics of householders and family members, including their education, health, trust, family income, and family loans. The database has a large sample size and strong credibility. Based on this database, there have been many good studies.

#### *4.2. Empirical Model*

#### 4.2.1. Ordered Probit Model

The data reflecting the credit availability of farmers are ordered variables of classification, and there are three categories, including "weak credit availability", "general credit availability", and "strong credit availability". Therefore, we use an ordered probit model for the estimation. The model is set as follows:

$$Y\star = X'\beta + \varepsilon \tag{1}$$

$$Y = \begin{cases} 1, & \text{Y} \preceq \chi\_0 \\ 2, & \text{Y}\_0 < \text{Y} \preceq \chi\_1 \\ 3, & \text{Y} > \text{Y}\_1 \end{cases} \tag{2}$$

where *Y\** is an unobservable latent variable, and *Y* is the observation-dependent variable. *Y* = 1 means that the farmer's loan application is rejected; that is, the credit availability of farmers is weak. *Y* = 2 means that the farmer's loan application is accepted but not fully satisfied; that is, the credit availability of farmers is general. *Y* = 3 means that the farmers' loan applications are fully satisfied; that is, the credit availability of farmers is strong. *X* is the set of explanatory variables, which may affect the credit availability of farmers. γ<sup>0</sup> and γ<sup>1</sup> are unknown cutoff points, and satisfy γ0<γ1. We assume ε~N (0, 1), the probability of variable Y taking each value is:

$$P(Y=1|X|\text{)} = P(Y \ast \le \gamma\_0 | X) = P(X^\prime \beta + \varepsilon \le \gamma\_0 | X) = P(\varepsilon \le \gamma\_0 - X^\prime \beta | X) = \phi(\gamma\_0 - X^\prime \beta) \tag{3}$$

$$P(\mathbf{Y} = \mathbf{2}|\mathbf{X}) = P(\gamma\_0 < \mathbf{Y} \star \le \gamma\_1 | \mathbf{X}) = P(\mathbf{Y} \star \le \gamma\_1 | \mathbf{X}) - P(\mathbf{Y} \star \le \gamma\_0 | \mathbf{X}) = \phi(\gamma\_1 - X'\beta) - \phi(\gamma\_0 - X'\beta) \tag{4}$$

$$P(Y = \Im |X|) = P(Y \star \gamma \gamma\_1 | X) = 1 - P(Y \star \le \gamma\_1 | X) = 1 - \phi(\gamma\_1 - X^\prime \beta) \tag{5}$$

Formula (3) refers to the probability that a farmer belongs to the group with a weak credit availability under the influence of factor X. Formula (4) refers to the probability that a farmer belongs to the group with a general credit availability under the influence of factor X. Formula (5) refers to the probability that a farmer belongs to the group with a strong credit availability under the influence of factor X. We use maximum likelihood estimation for testing. It uses a probability model to maximize the probability of the observed sample data. Then, the loglikelihood function will be:

$$\begin{split} \operatorname{Ln} \operatorname{Ln} \{ \beta\_{\prime} &\quad \gamma\_{0}, \gamma\_{1} \} &= \operatorname{Ln} [\operatorname{P}(Y = 1 | \mathbf{X} ) \bullet \operatorname{P}(Y = 2 | \mathbf{X} ) \bullet \operatorname{P}(Y = 3 | \mathbf{X} )] \\ &= \operatorname{Ln} \, \phi(\gamma\_{0} - \mathbf{X}^{\prime} \beta) + \operatorname{Ln} \, \left[ \phi(\gamma\_{1} - \mathbf{X}^{\prime} \beta) - \phi(\gamma\_{0} - \mathbf{X}^{\prime} \beta) \right] + \operatorname{Ln} \, \left[ 1 - \phi(\gamma\_{1} - \mathbf{X}^{\prime} \beta) \right] \end{split} \tag{6}$$

Using the maximum likelihood estimation method, we can get the parameter β, γ0, and γ1, that is, the influence of the explanatory variables (X) on the credit availability of farmers (Y).

#### 4.2.2. Propensity Score Matching Model

The ordered probit regression model can only give us a general answer to the impact of human and social capitals on the credit availability of farmers. In particular, variables such as farmers' education, party membership, and cooperative membership, are all farmers' self-selections. These selections may

not be random, which may lead to some errors in the estimation of the ordered probit model. However, the propensity score matching model can effectively control the selection bias problem through a counterfactual estimation [41,42]. Therefore, for these important human and social capital variables, we used the propensity score matching model to test them again to obtain a more accurate result. Its basic principle is:

$$D\_i = \begin{cases} 1, & Z\_i\alpha + \mu\_i > 0 \\ 0, & Z\_i\alpha + \mu\_i \le 0 \end{cases} \tag{7}$$

$$Y\_{\dot{i}} = \begin{cases} \begin{array}{c} Y\_{1\dot{i}\dot{\prime}} \\ Y\_{0\dot{i}\dot{\prime}} \end{array} & \begin{array}{c} If \quad D\_{\dot{i}} = 1 \\ If \quad D\_{\dot{i}} = 0 \end{array} \tag{8}$$

where Z (Z , X) represents the factors affecting the choice of the farmer, *D<sup>i</sup>* = 1 represents the treatment group, *D<sup>i</sup>* = 0 represents the control group, *Y*1*<sup>i</sup>* represents the credit availability of the treatment group farmer *i*, and *Y*0*<sup>i</sup>* represents the credit availability of the control group farmer *i*. The problem of selection bias is that under the influence of certain factors (Z), farmers cannot randomly choose to enter the control group or the treatment group. This leads to the general model estimation results not completely excluding the influence of other factors and provides the net influence (*Y*1*<sup>i</sup>* − *Y*0*<sup>i</sup>* ) of the variable D.

Taking the cooperative membership of farmers as an example, for a farmer *i* participating in a cooperative organization (treatment group), *Y*1*<sup>i</sup>* means the credit availability of farmer *i*, and *Y*0*<sup>i</sup>* means the credit availability of farmer *i* if he does not participate in the cooperative organization. Then, the difference between the two (*Y*1*<sup>i</sup>* and *Y*0*<sup>i</sup>* ) is the net influence of the cooperative membership on the credit availability of farmer *i*. However, in fact, *Y*0*<sup>i</sup>* is not observable, so an approximate estimate of *Y*0*<sup>i</sup>* needs to be found to obtain the net influence of the cooperative membership.

The basic idea of the propensity score matching estimation is as follows. First, the propensity score of each farmer is obtained according to the logit regression, that is, the probability of a farmer entering the treatment group under the influence of factor Z.

Second, for farmer *i* in the treatment group, according to a certain matching method (e.g., kernel matching and nearest-neighbor matching), we find a farmer *j* in the control group whose propensity scores are as equal or close to farmer *i* as possible. Thus, we can assume that the probability of farmers *j* and *i* entering the treatment group are the same or similar. Then, we take the *Y*0*<sup>j</sup>* corresponding to farmer *j* as the matching estimator of *Y*0*<sup>i</sup>* , that is, *Y*ˆ <sup>0</sup>*<sup>i</sup>* = *Y*0*<sup>j</sup>* .

Finally, we can get the net influence (treatment effect) of a variable (D) on farmer *<sup>i</sup>*: *<sup>Y</sup>*1*<sup>i</sup>* <sup>−</sup> *<sup>Y</sup>*<sup>ˆ</sup> <sup>0</sup>*<sup>i</sup>* = *Y*1*<sup>i</sup>* − *Y*0*<sup>j</sup>* . The average treatment effect (ATT) of the treatment group is:

$$ATT = E\left(Y\_{1i} - \hat{Y}\_{0i}\right) = E\left(Y\_{1i} - Y\_{0i}|D\_i = 1\right) = E\left(Y\_1|D\_i = 1\right) - E\left(Y\_0|D\_i = 1\right) \tag{9}$$

#### *4.3. Variable Definition and Descriptive Statistical Analysis*

*Dependent variable.* The measurement of the farmers' credit availability was mainly based on the question, "Has there been any situation where your family's borrowing application was rejected or the borrowing amount obtained was less than the amount of the application?" This can reflect the degree of the farmers' credit availability. If the answer was "all borrowing applications were fully satisfied", this means that the farmers' credit availability was strong, and the value was 3. If the answer was "all borrowing applications were accepted, but the loan obtained was less than the requested amount", this means that the farmer's credit availability was general, and the value was 2. If the answer was "one or more borrowing applications were rejected", this means that the farmer's credit availability was weak, and the value was 1. We divided the credit availability of farmers into formal and informal credit availabilities. We referred to farmers' borrowing from banks, rural credit cooperatives, and other formal credit organizations as formal credit. Then, we referred to farmers' borrowing from relatives or friends as informal credit.

*Independent variables.* The independent variables selected in this study mainly included the characteristics of farmers' human and social capitals, as well as the personal and family characteristics of the householders.

*Human capital*. We mainly selected two variables to reflect the human capital of farmers: the education and health of householders. Based on the division method of the agricultural census in China, we divided the education level into values of 1–5, representing "never attended school," "primary school", "middle school", "high school", and "university", respectively. The measurement of the farmers' health was mainly based on the self-evaluation of farmers. The value was 1–5, representing "worse health", "bad health", "general health", "good health", and "better health", respectively.

*Social capital*. We divided the social capital of farmers into political relationship capital, organizational relationship capital, interpersonal relationship capital, and financial relationship capital. Among them, political relationship capital is mainly expressed by the political status of farmers, including the party membership and cadre status of farmers. Interpersonal relationship capital mainly used three questions: "How many brothers and sisters do you have?" "Do you think relatives and friends can be trusted?" and "Do you think anyone other than relatives and friends can be trusted?" The options of the latter two questions included "very untrusted", "not very trusted", "generally trusted", "relatively trusted", and "very trusted", which were assigned values of 1–5 respectively. Organizational relationship capital was mainly expressed by participation in cooperative economic organizations. Financial relationship capital was mainly expressed by the work industry of family members. If there were family members working in the financial industry, then the value was 1. If there were no family members working in the financial industry, then the value was 0.

*Personal and family characteristics of householders*. Referring to previous research, we selected householders' age, gender, marital status, outside working experience, family size, family per-capita income, and family wealth as the control variables.

In the estimation, the direct introduction of categorical variables may result in an inaccurate coefficient estimation and economic meaning. Therefore, for the four categorical variables—education, health, the trust of relatives and friends, and the trust of other people—we set eight dummy variables to make the estimation results more accurate, including "education1", "education2", "health1", "health2", "rela\_trust1", "rela\_trust2", "others\_trust1", and "others\_trust2". The definitions and descriptive statistics for each variable are shown in Table 1.


#### **Table 1.**Descriptive statistics of the selected variables.

### **5. Empirical Analysis**

First, we used the ordered probit model to estimate the impact of human and social capitals on the formal and informal credit availability of farmers (see Table 4). Second, for some human and social capital variables with obvious self-selectivity, we made a counterfactual estimation with the propensity score matching model—a bias-corrected matching estimation. This could reduce the sample selection bias and provide the average treatment effect on the formal and informal credit availability of farmers. The results are shown in Table 5.

#### *5.1. Sample Basic Characterization*

The basic characteristics of sample farmers are shown in Table 2. The gender of householders is mainly male, accounting for 91.6% of the total sample. The age of the householders is between 21 and 101 years. Householders aged 40–60 years old account for 64.3% of the total sample. Most of the householders have a low education level; about 80.5% of these farmers have a primary or middle-high-school education level. The per capita income of most households (80.9%) is below CN¥ 15,000. These statistical characteristics (e.g., low education and low income) are roughly consistent with the basic situation of rural households in China. Consequently, the sample selection had a certain credibility.


**Table 2.** The basic characteristics of the surveyed farmers.

Table 3 shows that about 2792 sample farmers chose informal borrowing channels; the total rate was 89.3%. About 1036 sample farmers chose formal borrowing channels; the total rate was 33.1%. This indicates that farmers prefer to choose informal borrowing channels. The sum of farmers using formal and informal credit channels is greater than the total sample, because some farmers applied for both formal and informal borrowing (about 701 farmers).


Table 3 also shows the credit availability of farmers. About 69% of the sample farmers' borrowing applications were fully satisfied, about 16% of the sample farmers' borrowing applications were accepted but not fully met, and about 15% of the sample farmers' applications were rejected one or more times. This means that about 31% of the sample farmers' applications could not be fully met. There is still room for improvement in rural financial development in China. Furthermore, compared with informal credit, the probability of farmers being rejected by formal financial institutions increased significantly (14.7% > 12.8%).

#### *5.2. Estimation Results*

#### 5.2.1. Ordered Probit Model Estimation Results

Some studies show that social capital had a significant impact on the health of residents [43]. In order to reduce the impact of multicollinearity between human and social capitals on the estimation results, we referred to the treatment of Ding et al. [44]. In the estimation, the model (2) and (5) only included human capital variables and control variables, and the models (3) and (6) only included the social capital variables and control variables. The estimation results are shown in Table 4. We can see from this result that R2 is very low. An important reason for the lower R2 is missing variables. The credit availability of farmers is not only related to the characteristics of the farmers themselves, but also to the characteristics of the lenders. In the model, we mainly examine the characteristics of the farmers themselves, and lack the characteristic data of the corresponding lenders. This may result in a lower R2 in our model. In addition, there are generally two purposes for using regression models. One is explanation, and the other is prediction. If you need accurate predictions, a lower R2 is not feasible. However, if you want to analyze the significance of the explanatory variables, a lower R2 is feasible. The purpose of our model is mainly to explain, not to predict. However, the R2 of the ordered probit model is low, and there may be some deviation in the parameter estimation. Thus, we used the PSM model to estimate again. We discussed the same conclusions of the two models in the result section of the ordered probit model. We discussed different conclusions after the PSM estimate.

The results showed that human capital had a positive impact on the formal credit availability of farmers, but that it had no significant impact on the informal credit availability of farmers (Table 4). Specifically, for formal credit, compared with farmers with bad or worse health (the control group), the dummy variables "health1" and "health2" had a significant positive impact on the formal credit availability of farmers. This indicates that improving farmers' health will play an important role in increasing the formal credit availability of farmers. This is consistent with our hypothesis. Healthier farmers have higher incomes and a stronger ability to make continuous payments, reducing the loan risk of banks. Compared with farmers with a primary or lower education level, the dummy variables "education1" and "education2" had a positive but insignificant impact on the formal credit availability of farmers. This is inconsistent with our hypothesis and the results of most scholars. It may be because the investment of farmers in education is influenced by many factors, such as family income. Some scholars believe that education is also related to the credit situation of families, and farmers with credit constraints always have a lower investment in education [45]. Therefore, there may be a degree of causality between education and family credit, causing endogenous problems. We will use the more accurate method of PSM for testing before discussing this further. For informal credit, only the dummy variable "health2" had a significant impact on the informal credit availability of farmers at the level of 10%, and the dummy variables "health1", "education1", and "education2" had no significant impact on the informal credit availability of farmers. As the saying goes, "birds of a feather flock together". People within the group have some similarities. Therefore, there may be little difference in the education level between friends. Borrowing between relatives and friends is based more on the social relationship formed by geographical and blood ties. However, the human capital of farmers is not particularly important to their relatives and friends.


**Table 4.** The estimation results of the ordered probit model.

Notes: (1) Standard errors in brackets, \*\*\* *p* < 0.01, \*\* *p* < 0.05, \* *p* < 0.1. (2) In the estimation, we removed the samples that answered "unclear" when measuring the trust of family and friends and of other people, so the number of observation values in the regression is not consistent with the observation values in the descriptive statistical analysis above.

The results also showed that social capital had a significant impact on the credit availability of farmers. Among them, interpersonal relationship capital had a significant impact on the formal

and informal credit availability of farmers. This is consistent with our hypothesis. However, the impacts of financial relationship capital on the formal and informal credit availability of farmers were all non-significant. This is inconsistent with our hypothesis and the conclusions of most previous authors. It may be because the financial relationship capital of our sample farmers is generally weak, and only a few farmers have relatives working in banks, resulting in the estimated results not being statistically significant.

Specifically, for formal credit, the estimation results of models (1)–(3) were relatively consistent. For political relationship capital, the variable "cadre" had a significant positive impact on the formal credit availability of farmers, while the variable "party membership" had a positive but insignificant impact on the formal credit availability of farmers. These results also support the findings of Xu and Yang [46]. This shows that political relationship capital has a positive impact on farmers' formal credit availability. For interpersonal relationship capital, the variables "siblings" and "others\_trust2" had a significant positive impact on the formal credit availability of farmers, while the variables "rela\_trust1" and "rela\_trust2" had a positive but insignificant impact on the formal credit availability of farmers. For organizational relationship capital, the impact of the variable "cooperative membership" on the formal credit availability of farmers was not significant. This may be because the development of farmer cooperative organizations is still not perfect in China. Many cooperative organizations are just in the form of cooperatives and have no substantive operations, which reduces the trust of formal financial institutions in cooperatives. Therefore, whether farmers participate in cooperative organizations or not has no significant impact on the formal credit availability of farmers [8].

For informal credit, the estimation results of models (4)–(6) were also relatively consistent. The variables "cooperative membership", "siblings", "rela\_trust2", "others\_trust1," and "others\_trust2" all had a significant positive impact on the informal credit availability of farmers. This shows that farmers with better organizational relationship capital and interpersonal relationship capital are more likely to obtain loans from their relatives and friends. However, the impact of political relationship capital on the informal credit availability of farmers was not significant. This is inconsistent with our hypothesis, and we will use the more accurate method of PSM for testing before discussing this further.

#### 5.2.2. Propensity Score Matching Estimation Results

The propensity score matching model can only deal with binary variables in general. However, the variables of farmers' education, trust in relatives and friends, and trust in other people are not binary variables. For such variables, some scholars point out that we can perform pairwise matching between groups, select one group at a time, and match with the rest of the groups one by one [47,48]. For the education variable, we took farmers with primary school or lower education as the control group, farmers with middle school education as one treatment group, and farmers with high school or above education as another treatment group. Then, we matched the two treatment groups with the control group one by one. Similarly, for the two trust variables, we took farmers who selected "very untrusted" and "not very trusted" as the control group, farmers who selected "generally trusted" as one treatment group, and farmers who selected "relatively trusted" and "very trusted" as another treatment group. Then, we also matched the two treatment groups with the control group one by one. There are many methods for propensity score matching. We used the bias-corrected matching estimator to measure the average treatment effect on the treated (SATT) variables of human and social capitals. The estimated results are shown in Table 5.


**Table 5.** The estimation results of the propensity score matching model.

Notes: In the matching of farmers' education, we took farmers with primary school or lower education as the control group. In the matching of the trust between relatives or friends, we took farmers who thought their relatives or friends were not trustworthy as the control group. Additionally, in the matching of the trust of other people, we took farmers who thought other people were not trustworthy as the control group. \*\*\* *p* < 0.01, \*\* *p* < 0.05, \* *p* < 0.1.

In terms of human capital, unlike the estimation results of the ordered probit model, the education level of farmers ("high school or above") had a significant positive impact on their formal credit availability. This indicates that, compared with farmers with primary school or lower education, farmers with high school or above education are more likely to obtain loans from formal financial institutions. This is inconsistent with the results of the ordered probit model but more consistent with our hypothesis. This may be because the PSM estimate excluded the effects of other factors and obtained the net impact of education on the credit availability of farmers. Our analysis is also mainly based on the estimated results of PSM. Farmers with higher education levels have a stronger credit consciousness, repayment ability, and willingness. They are more able to gain the trust of formal financial institutions. Other estimation results were consistent with the ordered probit model. Overall, based on the estimation results of the ordered probit and PSM models, human capital had a significant positive impact on the formal credit availability of farmers. However, its impact on the informal credit availability of farmers was not significant.

In terms of social capital, unlike the estimation results of the ordered probit model, the cadre status of farmers also had a significant positive impact on their informal credit availability at a 10% level. This indicates that the cadre status of farmers is also beneficial to them for obtaining loans from relatives and friends. This may be because rural cadres are generally elected by farmers and usually have a high prestige and credibility in the rural group. Other estimation results were consistent with the ordered probit model. Overall, based on the estimation results of the ordered probit and PSM models, the interpersonal relationship capital and political relationship capital of farmers had a significant positive impact on the formal and informal credit availability of farmers. The organizational relationship capital only had a significant positive impact on the informal credit availability of farmers. However, the impacts of financial relationship capital on the formal and informal credit availability of farmers were all insignificant, which may be related to our sample selection. Only a few farmers had relatives working in the financial sector. Therefore, the sample matching results may not have been ideal, and we do not comment on this here.

#### **6. Conclusions and Suggestions**

Based on the CHIP2013 database, we estimated the impact of human and social capitals on the formal and informal credit availability of farmers with the ordered probit model and the PSM model. The basic conclusions are as follows.

First, for the borrowing channels, farmers preferred to choose an informal channel. For the credit availability of farmers, about 31% of the sample farmers did not receive full loans, which is similar to the findings of Yu and Zhou (25.8%) [49]. This indicates that there is still room for further improvement of rural finance in China.

Second, human and social capitals both had a certain positive impact on the formal credit availability of farmers, while for informal credit the impact of social capital was more significant. The hypothesis that social capital has a significant positive effect on the formal and informal credit availability of farmers has been confirmed, but the hypothesis that human capital has a positive effect on the informal credit availability of farmers has not been confirmed. This may be because the informal borrowing we studied was mainly between relatives and friends, which is mainly based on the social relationship between them. The trust and prestige formed by the social relationship based on blood, kinship, and geography have a great influence on farmers' borrowing. In such a relationship, the impact of farmers' education or health is relatively weak. This shows that formal financial institutions not only pay attention to the human capital of farmers but also to their social capital to reduce the risk of lending, while informal lenders, that is, their relatives and friends, pay more attention to the social capital of farmers.

Third, specifically for human capital, the education and health level of farmers had a positive and significant impact on their formal credit availability. Farmers with a higher education level, especially with a high school or above education level, were more likely to obtain loans from formal financial institutions. For social capital, interpersonal relationship capital had a positive impact on the formal and informal credit availability of farmers. This indicates that more siblings and a higher trust among relatives, friends, and others can greatly help farmers obtain loans. Organizational relationship capital only had a significant positive impact on the informal credit availability of farmers. This may be because the mutual assistance and supervision of cooperatives make it easier for farmers to obtain loans from cooperative members, including relatives and friends. However, cooperative organizations are not fully recognized by formal financial institutions, and this has no significant impact on the formal credit availability of farmers. For political relationship capital, compared with the party membership of farmers, farmers' cadre status can give farmers greater prestige and better help them obtain loans. This may be because the cadre identity information of farmers is a kind of explicit information, while the party membership information of farmers is a kind of relatively implicit information. The behavior of village cadres is more constrained, and the cost of default is higher. However, the hypothesis that financial relationship capital has a positive impact on the formal and informal credit availability of farmers has not been confirmed. This may be because the financial relationship capital of farmers is generally weak. Fewer farmers have relatives working in banks. In other words, in the estimation process, the difference in independent variables is small, difficult to render statistically significant. Based on this, we do not overly discuss the impact of financial relationship capital on the credit availability of farmers.

Based on the above conclusions, we can state the following recommendations. For farmers, improving their human capital and social capital has a positive effect on their access to credit. For example, improving farmers' education level, health level, and their prestige, joining cooperative economic organizations, and strengthening a close relationship with relatives and friends have important positive effects on improving the credit availability of farmers. For the government, first, in rural finance more farmers prefer to choose informal borrowing channels, and the role of formal financial institutions needs to be further improved. Second, in terms of human capital, improving the health and education of farmers can help them obtain formal loans. In particular, popularizing high school or above education levels can enhance the formal credit availability of farmers. Third, regarding social capital, Dinh et al. [35] argued that it is difficult to put forward an effective policy recommendation to the government or banks that generally fosters social capital. Moreover, such policy measures may have some unexpected and unnecessary side effects. However, we contend that improving the participation of farmer cooperative organizations and their social recognition may be an important and effective way to improve the credit availability of farmers. Only through the joint efforts of the government and farmers themselves, can we effectively improve the credit availability of farmers.

We have studied the impact of farmers' education and health (especially health) on the credit availability of farmers. Our study improved the research framework of human capital on farmers' credit availability to a certain extent. However, whether there is an interaction between human capital

and social capital, and how the interaction between them affects the credit availability of farmers, also has some research value.

**Author Contributions:** Conceptualization, J.L.; methodology, J.L. and G.Z.; data curation, J.L. and G.Z.; writing-original draft preparation, J.L.; writing-review and editing, G.Z. and J.Z.; funding acquisition, C.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by National Natural Science Foundation of China: 71673103. Humanities and Social Science Project of the Department of Education of Hubei Province: 18Q138.

**Acknowledgments:** The authors would like to thank Yangmin Ke for his comments and help, as well as the reviewers and editor for their constructive comments. Any errors and misinterpretations remain the authors'.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **The Role of External Support on the Implementation of Resource Efficiency Actions: Evidence from European Manufacturing Firms**

**David Aristei and Manuela Gallo \***

Department of Economics, University of Perugia, 06123 Perugia, Italy; david.aristeil@unipg.it **\*** Correspondence: manuela.gallo@unipg.it

**Abstract:** This paper contributes to analyze the relationship between firms' recourse to different types of external support and adoption of environmental sustainability practices. To this aim, we consider both direct financial support and indirect support, in the form of advice and consulting services, upon which the firm relies on in its efforts to be more resource efficient. The empirical analysis uses data on 6595 manufacturing firms from 35 European countries, taken from the third and fourth waves of the Flash Eurobarometer survey "Small and Medium Enterprises, Resource Efficiency and Green Markets". Our empirical findings suggest that firms using external financing and external advice are more likely to implement greening investments and practices. Moreover, we provide strong empirical evidence that external support significantly contributes to increase the benefits from the adoption of resource efficiency actions in terms of production cost reduction. This study further contributes to the existing literature by highlighting the heterogenous effects of direct and indirect external support on the environmental sustainability actions of both SMEs and large firms.

**Keywords:** external support; environmental practices; resource efficiency; sustainable entrepreneurship; firm size

#### **1. Introduction**

In recent years, increasing attention has been paid to the adoption of resource and energy efficiency practices by firms, also as a result of government policies aimed at supporting the implementation of environmental technologies and eco-innovations. Energy and resource efficiency is a key pillar of the European Union's long-term strategic vision and it is part of the Sustainable Development Goals approved by the United Nations in the Agenda 2030, with the aim to encourage a profound systemic shift to a more sustainable economy.

Firms' adoption of measures for a more responsible and efficient use of natural resources is important both for overcoming the problem of resource scarcity and waste management, and for incentivizing sustainable development and innovation towards a circular economy [1]. The increase in energy consumption, the need to reduce emission of greenhouse gases, the progressive depletion of natural resources and the dependence on energy from countries characterized by unstable political regimes have generated the need for eco-innovative solutions. It follows that firms are called upon to change their business model, taking into greater account the environmental and social values.

Firms' ability to integrate and align multiple forms of value (commercial, economic and financial values and environmental and social values) within their business models is a timely and important issue that affects not only firms' development but even the economic system as a whole. The challenges in this field are numerous, especially for small and medium enterprises (SMEs). This new development model is based on the resource efficiency in a logic of circularity, and on innovation in terms of eco-design, with the aim to reduce negative externalities on the environment since the design and development phases.

**Citation:** Aristei, D.; Gallo, M. The Role of External Support on the Implementation of Resource Efficiency Actions: Evidence from European Manufacturing Firms. *Sustainability* **2021**, *13*, 9531. https: //doi.org/10.3390/su13179531

Academic Editor: Alan Randall

Received: 30 July 2021 Accepted: 19 August 2021 Published: 24 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The general benefits deriving from eco-sustainable practices are diversified according to the characteristics of the intervention, the degree of circularity of production processes, the external environment in which the company operates, and the role of the company in the value chain. Currently, the most common resource efficiency measures among firms include: the environmental management systems aimed at saving water, energy and greenhouse gas emissions; the prevalent use of energy from renewable sources; the minimization of waste and sale of waste materials to another company; the reuse of materials.

Environmental investments require a strategic long-term vision: they do not represent exclusively a cost item for companies, but a central factor in acquiring a competitive advantage in the medium and long term, a better reputation and different strategic positioning. The European Commission has repeatedly stressed that in order to support firms in developing new resource-efficient technologies and solutions, it is crucial for access to financing; nevertheless, the context of technological and market uncertainties related to the eco-innovations may contribute to increase the difficulty in access to external funding by firms. Moreover, to attract investment in eco-innovations can be difficult, as they are often characterized by high risk and long-term returns [2–4].

There are several factors that may hamper or delay the adoption of resource and energy efficiency measures [5]. The literature has documented the existence of direct negative effects of financial barriers on the adoption of resource and energy efficiency practices [6,7]. For this reason, many targeted public financing programs have been proposed by both EU and non-EU countries, and a large set of financial instruments and services have been designed by financial intermediaries to support firms' investments. However, only in a few cases has the link between different types of external support and the adoption of resource efficiency practices been analyzed [8].

In this paper, we shed light on the relationship between the use of different types of external support and the adoption of eco-innovative practices by small and medium enterprises. To this aim, extending the analysis of Bodas-Freitas and Corrocher [8], we consider both direct financial support and indirect support in the form of advice and consulting services related to the adoption of resource efficiency practices. Furthermore, as a novel contribution to the literature, in this study, we assess the presence of firm-size heterogeneities in the effects of direct and indirect support on firms' engagement in the greening process. In fact, while the environmental behavior of small- and medium-sized enterprises has been extensively analyzed, that of large firms is still broadly unexplored. Our analysis relies on cross-sectional data from the third and fourth waves of the Flash Eurobarometer surveys "Small and Medium Enterprises, Resource Efficiency and Green Markets" [9,10]. In particular, our estimation sample consists of 6595 manufacturing firms from 35 European countries. The empirical analysis is carried out by means of a propensity score matching approach, which compares the effect of the treatment (in our case, the use of external support) between the subsamples of treated and untreated firms with similar observable characteristics. This methodology allows us to produce an estimate of the average additional effect of external support on the probability of adopting resource efficiency practices and on the benefits from the adoption, the so-called average treatment effect on the treated (ATET).

Our main results suggest that firms using external financial advice are more likely to implement greening investments and practices, including re-engineering and waste management actions. Moreover, we provide strong empirical evidence that external support significantly contributes to increase the benefits from the adoption of resource efficiency measures in terms of production cost reduction. This study further contributes to the existing literature by providing strong empirical evidence of the heterogeneous effects exerted by direct and indirect external support on different types of environmental practices of both SMEs and large firms. In this latter respect, our findings suggest that while SMEs' implementation of environmental practices strongly depends on both external financing and indirect support, large firms tend to rely more on internal financial resources and

mainly benefit from the recourse to external advice and consulting services to support their efforts to be more resource efficient.

The paper is structured as follows. Section 2 provides an overview of the background literature. Section 3 describes the data, while Section 4 illustrates the methodology. Section 5 presents and discusses the main results and Section 6 offers some concluding remarks and policy implications.

#### **2. Literature Review**

In the literature, there are many attempts to develop a taxonomy of the barriers to investments in energy and resource efficiency practices. Sorrell et al. [11] classify these barriers according to economic, organizational, or behavioral categories, observing that they may co-exist and overlap each other. Rentschler et al. [12] propose a taxonomy consisting in the following five categories: scarce information, low capacity, financial constraints, uncompetitive market structures and fiscal mismanagement. In particular, they observe how the theoretical assumptions of perfect and efficient markets are violated in practice, and how this results in investment barriers. Analogously, Jordan et al. [13] demonstrate that deficits in innovation culture, inter-firm cooperation along the value chain, finance, awareness and take-up of government funds limit the adoption of resource efficiency measures. They propose a policy mix, comprising government funding schemes, innovation agents and innovation laboratories, to support firms in implementing resource efficiency procedures.

Previous empirical studies have also focused attention on the role of external support on the adoption of different environmental practices. Bodas-Freitas and Corrocher [8] assess the impact of both direct financial support and indirect external support, in term of advice and consultancy, on the adoption of resource efficiency practices. They find that both types of external support positively affect the firm's implementation of resource efficiency measures and the cost-reduction benefits of adopting resource efficiency practices. They also point out that financial support has a direct effect on the benefits from the adoption, while the recourse to advice and consultancy support affects them indirectly, by supporting the implementation of complementary technological and managerial solutions. Accordingly, Hoogendoorn et al. [14] show that companies that receive external financial support are not only more likely to invest in practices related to production processes (greening processes), but they are also more likely to offer green products and services.

Firms recognize limited access to capital as one of the most common barriers to resource efficiency [12,15]. In particular, small and medium-sized enterprises experience greater difficulties in access to financing. These firms are typically characterized by scarce own funds and usually rely on bank loans than equity to finance their activity; moreover, the lack of collateral and the dependence by few clients may contribute to the perception of vulnerability and risk by financial intermediaries and limit the probability of advantageous funding conditions [16]. Furthermore, investments in eco-innovations and resource efficiency are characterized by high technical risks and longer-term returns, hence financial constraints are particularly relevant for these kinds of investments and may further contribute to affect firms' probability of experiencing liquidity constraints [2,17]. The adoption of energy and resource efficiency measures, even when it represents a profitable investment, has a crucial barrier in the lack of access to capital. Anderson and Newell [18] and Thollander et al. [19], among others, conclude that the initial investment costs negatively affect the adoption rate, especially for larger investments. Moreover, during periods of banking sector instability, as, for example, financial and economic crises, the restricted access to credit may further contribute to limit firms' resource efficiency investments [12].

To overcome financial constraints and capital–market imperfections, many public policy measures have been introduced with the aim to complement inefficient credit markets and provide SMEs with financial incentives and assistance supporting the innovation process [13,20]. Investment subsidies or soft loans can contribute to the dissemination of energy-efficiency measures in SMEs [7] and can increase access to finance for eco-innovative

activities. Access to public funds and incentives is considered effective for improving a firm's ability to introduce eco-innovations. In this respect, Ghisetti [21] shows the crucial role of governmental demand in shaping the direction and speed of environmental innovations in the manufacturing sector. Accordingly, Özbu ˘gday et al. [22] provide evidence of a positive and statistically significant effect of resource efficiency investments on SMEs' growth performances and suggest that an effective policy that governments could adopt to boost green growth is to give public subsidies to support resource efficiency investments of SMEs operating in energy-intensive sectors.

According with the previous remarks, we thus posit our first research hypothesis:

**Hypothesis 1** (**H1)**. *Access to private and public external financing enhances firms' engagement in different types of environmental practices and improves the benefits from the adoption of greening processes.*

Both banks and institutional investors, as well as policy makers, thus exert a crucial role in mobilizing a large amount of funds and allocating them to long-term environmental or eco-innovative projects that are often characterized by immature technologies or complex technological systems [23]. On the other hand, it is worth remarking that the extent to which banks appreciate the potential profitability of resource efficiency projects also depend on a firm's ability to report efficiency practices and communicate the related opportunities [24]. In this sense, the lack of information on these technologies or the lack of specific expertise by firms are further important obstacles that may hamper their access to financial resources. Previous studies point out that more than half of the European SMEs recognize information constraints as an obstacle to improving resource efficiency [9,10,25] and that smaller enterprises have a greater perception of the barriers to energy efficiency than larger ones, discouraging them from adopting energy-efficient technologies and practices [26]. In addition to the knowledge gaps, SMEs' capabilities to implement new measures of resource and energy efficiency may be constrained by the lack of time, human capital, managerial/organizational factors, and informal management of sustainability issues [16,27]. Thus, external advice and consultancy may provide firms with the competencies necessary to implement resource efficiency actions, enhance the efficiency of these measures and allow sustainable innovations [28]. Accordingly, Horbach et al. [29] show that eco-innovative activities require more external sources of knowledge and information compared to other types of innovations; moreover, they confirm the central role of regulation and cost savings as motivations for eco-innovations. In addition to these knowledge gaps, SMEs are typically characterized by organizational rigidity that act as barriers to firm performance and have a significant impact on innovation capability [30]. Moreover, firms with strong business networks and easy access to knowledge and technology are more likely to conduct eco-innovation activities [23].

The pursuit of green growth requires direct financial investments and indirect forms of external support, involving both private and public actors. In particular, consultancy or other advice are aimed at filling organizational, knowledge and technical gaps, that can be provided by suppliers, consulting firms or public research institutions. Thus, while financial barriers result into low capital availability and low access to external funding opportunities and contribute to hamper firms' innovation and growth, the difficulty of identifying cost-effective resource efficiency projects represents an additional limit to the implementation of resource efficiency practices by firms. Thus, the recourse to both external funding and advice may significantly mitigate issues related to the scarcity of financial resources and the lack of expertise and knowledge, enhancing SMEs' ability to implement technologies and practices that can achieve savings in resources and production costs [8,29]. As observed by Ghisetti et al. [2], small firms face major difficulties in getting credit for their eco-innovation investments compared to large enterprises, which may have direct access to capital markets and have more developed skills and competencies to engage in environmental practices [14].

While the role of different types of external support on the environmental behavior of small and medium firms has been extensively analyzed, the effect on the greening processes of large enterprises still remains largely unexplored. Smaller firms are usually considered as lacking skills, knowledge and financial resources to implement environmental management systems [31,32], while larger enterprises are assumed to be more resourceful and proactive and so are more capable of enjoying the benefits deriving from the implementation of resource efficiency practices [33]. In this respect, openness to external sources of knowledge which could help small and medium firms to overcome the lack of internal capabilities and resources for the adoption of environmental activities [2]. Differences in strategic resource allocation patterns between small and large firms are also found to be related to firms' characteristics and organizational costs [34]. Moreover, Kalar et al. [35] pointed out that firms in the innovative stage of their organizational life-cycle are not only characterized by different resource efficiency strategies, but also have different external support needs than firms in the conservative stage. A further input for the environmental behavior of SMEs can derive from the "stakeholder perspective", according to which smaller firms are much more responsive to external pressure by stakeholders than large firms [14]. The relationship between stakeholder pressures and environmental strategy is found to vary with firm size [36]. In particular, SMEs pay more attention to achieve and maintain a good reputation, while large firms are better able to manage external pressures. On the other hand, as highlighted by Wong et al. [33], in many countries, large firms are often state-owned or subsidized by the government, so they could face governmental pressures for their involvement in environmental management.

Based on the discussion above, we posit and test the following two hypotheses:

**Hypothesis 2** (**H2)**. *Indirect external support, in the form of advice and consulting, facilitates the adoption of resource efficiency measures and increases the cost-reduction benefits from the adoption of such actions.*

**Hypothesis 3** (**H3)**. *The impact of alternative forms of external support on the extent and types of environmental practices is characterized by significant heterogeneity with respect to firm size.*

#### **3. Data and Descriptive Analysis**

#### *3.1. Data Sources*

To investigate firms' engagement in environmental practices, we rely on data from the Flash Eurobarometer survey "SMEs, resource efficiency and green markets" [9,10]. This survey is focused on SMEs and large firms operating in the Manufacturing, Retail, Services and Industry sectors in the 28 European Union Member States and other European and non-European countries. The survey provides detailed information on firms' investment and implementation of resource efficiency practices and use of different types of external support to introduce these measures, together with data on several firm-level characteristics. For the aims of our study, we combine cross-sectional data from two independent waves of the survey (the third and the fourth, referred, respectively, to years 2015 and 2017) and focus on manufacturing firms from 35 European countries (including both EU and non-EU states). The final estimation sample consists of a total of 6595 enterprises.

#### *3.2. Outcome Variables*

To assess firms' adoption of green processes, we consider several alternative measures. We first define a set of binary variables indicating which actions the firm is undertaking to be more resource efficient. In particular, we define 9 dummies equal to 1 if the firm has adopted actions to save water, save energy, use predominantly renewable energy, save materials, minimize waste, sell scrap material, recycle by reusing material or waste, design products that are easier to maintain or repair, and other resource efficiency measures. As in Bodas-Freitas and Corrocher [8], we sum these dummies to define alternative variables that count the number of resource efficiency practices implemented by the enterprise. We

first define a variable (*Resource Efficiency Actions*) counting the number of any resource efficiency measure adopted and taking values from 0 to 9. Then, we build two variables accounting for the different type of practices and define the variable *Re-engineering Actions* that counts the number of practices requiring process re-engineering (i.e., saving water, saving energy, using renewable energy, saving materials, and designing products easier to maintain/repair) and the variable *Waste Management Actions* that counts the number of practices related to waste management (i.e., minimizing waste, selling scrap material, recycling). As it can be noticed form Table 1, the number of resource efficiency actions implemented by the enterprises increases with firm size, according to the consideration that large firms have more resources to invest in greening processes. In particular, the number of re-engineering actions is equal to 2.459 in the subsample of SMEs, rising to 3.030 in that of large firms; analogously, the number of waste management actions varies from 1.693 for small and medium enterprises to 2.103 for large ones.


**Notes**: The table reports average values of the outcome, treatment and independent variables computed on the whole sample and on the subsamples of SMEs and large firms. Descriptive statistics are computed sample weights. **Source**: Own elaboration on Eurobarometer data.

We also consider firms' investment in resource efficiency as an additional proxy of involvement in greening processes and define a binary indicator (*Resource Efficiency Investment*) identifying firms that, over the last two years, invested at least 1% of their annual turnover to be more resource efficient. Even in this case, larger firms show a higher average value of the *Resource Efficiency Investment* indicator (0.486) with respect to small and medium ones (0.461), demonstrating that large firms have a greater propensity to invest part of their turnover in greening processes.

Finally, in order to assess the benefits from the adoption of resource efficiency practices, we define the dummy variable *Production Costs Decreased*, which is equal to 1 if the undertaken resource efficiency actions have contributed to decrease production costs over the past two years. This variable highlights that those firms that have higher environmental awareness also experience higher cost savings. Coherently with previous results, large firms experience larger savings in production costs (0.704), while SMEs have a slightly lower value (0.530). In any case, it is evident how the adoption of energy efficiency practices produce significant benefits in terms of reduction of production costs.

#### *3.3. Treatment Variables*

In our empirical analysis, we aim at investigating the role of the external support in the form of financial funding and business/technology advice on different types of external support on firms' implementation of resource efficiency practices and on the benefits derived from the adoption of greening processes. Following Hoogendorm et al. [14] and Bodas-Freitas and Corrocher [8], we consider alternative proxies for a firm's recourse to external support for implementing resource efficiency measures. First, we define a dichotomous variable (*Any External Support*) indicating whether the firm has relied on any form of external support in its efforts to be more resource efficient. Then, we define two additional dummies to distinguish between the use of (public and private) external financial support (*External Funding*) and the use of (public and private) advice and consultancy (*External Advice*). Finally, we define a binary variable (*External Funding and Advice*) identifying those firms that use both types of external supports, in order to assess the presence of possible complementarities on the effect of external funding and consultancy on the adoption of resource efficiency measures.

From Table 1, we notice that 27% of the firms in our sample use any form of external support; this percentage is equal to 25% for SMEs and increases to about 45% for large enterprises. External advice is the type of support mainly used by both SMEs and large firms, while only 6% of SMEs and nearly 12% of large firms rely on the combined use of funding and advice. The differences in average values of the resource efficiency indicators by support type are presented in Table 2. These values are always positive and statistically significant, with the only exception of the residual category *Other RE Actions*, demonstrating that the positive impact of external support on the implementation of greening actions. The practices requiring process re-engineering present the highest values: they range from 43%, when we consider any type of external support to 73% when the firm uses both external funding and advice. In particular, the combined use of direct and indirect external support (column 4) highlights the greatest effect on all the types of resource efficiency actions. When we consider investments in resource efficiency practices, we highlight that the highest impact is exerted by access to external funding, demonstrating that this type of external support significantly contributes to boost firm investments in greening practices. Finally, the effect of the different types of external support on the reduction of production costs varies from 7.7 to 15.1%, with the highest impact exerted by the combined use of external financing and advice. These results demonstrate that the use of any type of external support contributes to the increasing benefits for firms and could represent an incentive for the adoption of greening actions.


**Table 2.** Differences in average values of resource efficiency indicators by external support.

**Notes**: The table reports (unconditional) differences in the means/proportions of the outcome variables between the subsamples of firms recurring and not recurring to the different types of external support measures considered. \*\*\*, \*\* and \* denote significance of the differences in means/proportions at the 1, 5 and 10% levels, respectively. **Source**: Own elaboration on Eurobarometer data.

#### *3.4. Independent Variables*

As in Hoogendorm et al. [14] and Bodas-Freitas and Corrocher [8], we control for observable firm-level characteristics that might affect firms' decision to adopt environmental practices and to recur to external support. First, we control for firm age (in years) and size using binary indicators for *Small* (with 10–49 employees), *Medium* (with 50–249 employees) and *Large* (with 250 or more employees) enterprises (considering *Micro* firms with less than 10 employees as reference group). We also control for firm turnover by means of a dummy indicating firms with a turnover lower than 2 million Euro (*Low turnover*). Second, as a firm's decision to adopt resource efficiency measures depends not only on the external support received, but also on internal funds and competencies, we define two dummies (*Own financial resources* and *Own technical competencies*) to control whether the firm relies on its own financial resources and its own technical expertise to implement greening processes. Third, we control for the firm's market segment by means of three non-exclusive binary variables indicating whether the firm sells products or services directly to consumers (*B2C Market*), to other companies (*B2B Market*) and to public administrations (*PA Market*). Finally, we include survey year and country fixed effects, to control for heterogeneities in environmental practices and recourse to external support over time and across countries. Table A1 in the Appendix A reports complete variable definitions.

#### **4. Methods**

To assess the impact of external support on the firm's implementation of greening processes, we use a propensity score matching approach, which compares the effect of the treatment (i.e., the recourse to external support) in the subsamples of treated and untreated firms with similar observable characteristics. In particular, following previous literature [8,21,22], we focus on the additional effect of the external support on the adoption of resource efficiency actions and use Kernel matching algorithms to estimate the average treatment effect on the treated (ATET). Formally, the ATET can be written as:

$$ATET = E(\mathbf{Y}\_1 - \mathbf{Y}\_0 | D = 1) = E(\mathbf{Y}\_1 | D = 1) - E(\mathbf{Y}\_0 | D = 1) \tag{1}$$

where *D* = 1 indicates a firm's recourse to external funding and/or advice in its effort to be more resource efficient (treatment variable) and *Y*<sup>1</sup> and *Y*<sup>0</sup> represent the potential outcomes with and without treatment, i.e., the greening processes adopted by those firms that have recurred or have not recurred to external support, respectively. The ATET is the average treatment effect (*ATE* = *E*(*Y*<sup>1</sup> − *Y*0) = *E*(*Y*1) − *E*(*Y*0)) computed on the subsample of treated units. The last equality in Equation (1) highlights the counterfactual nature of a causal effect. The first term, *E*(*Y*1|*D* = 1), is the average outcome of treated units, which is an observable quantity. The second term, *E*(*Y*0|*D* = 1), refers instead to the average outcome of treated units had they not been treated; this quantity cannot be observed and a proper substitute for it has to be chosen in order to estimate the ATET.

When selection to treatment is not random (as it is in the case of a firm's decision to recur to external support), the treatment *D* is not probabilistically independent from *Y*<sup>1</sup> and *Y*0, giving rise to a selection bias and preventing proper identification of treatment effects. However, it is still possible to identify causal effects from observational data by assuming that the non-random assignment to treatment is driven by individual observable factors **x**. Under selection on observables, the knowledge of **x** may be sufficient to identify the causal parameters, even in a case of non-random assignment. In particular, as discussed in Rosenbaum and Rubin [37], the condition of randomization is restored by means of the so-called *Conditional Independence Assumption* (CIA), stating that, conditional on **x**, *Y*<sup>1</sup> and *Y*<sup>0</sup> are probabilistically independent of *D*: (*Y*1;*Y*0)⊥*D*|**x**. When the interest is in measuring average effects, it is possible to rely on a weaker assumption, the so-called *Conditional Mean Independence* (CMI), assuming that *E*(*Y*1|*D*, **x**) = *E*(*Y*1|**x**) and *E*(*Y*0|**x**, *D*) = *E*(*Y*0|**x**). Assuming CMI, we obtain that:

$$ATET(\mathbf{x}) = E(\mathbf{Y}\_1|\mathbf{x}, D=1) - E(\mathbf{Y}\_0|\mathbf{x}, D=1,) = E(\mathbf{Y}|\mathbf{x}, D=1) - E(\mathbf{Y}|\mathbf{x}, D=0) \tag{2}$$

which shows that, by conditioning on **x**, the *ATET*(**x**) depends on observable quantities and it is thus correctly identified and no bias emerges. By averaging *ATET*(**x**) over the support of **x**, we can then obtain the global effect:

$$ATET = E\_\mathbf{x} \{ ATET(\mathbf{x}) \} \tag{3}$$

implying that an estimation of the ATET can be obtained by the sample equivalent:

$$AT\hat{\mathbf{E}}T = \frac{1}{\sum\_{i=1}^{N} D\_i} \left\{ \sum\_{i=1}^{N} D\_i [\mathfrak{m}\_1(\mathbf{x}\_i) - \mathfrak{m}\_0(\mathbf{x}\_i)] \right\} \tag{4}$$

where *m*ˆ <sup>1</sup>(**x***i*) and *m*ˆ <sup>0</sup>(**x***i*) are consistent estimators of *E*(*Y*|**x**, *D* = 1) and *E*(*Y*|**x**, *D* = 0), respectively. Besides CMI, identification of average treatment effects (ATEs) also requires the *Overlap Assumption* (OA), which states that, for each unit, the probability to get treated given **x** (i.e., the *propensity score*) must be 0 < *P*(*D* = 1|**x***i*) < 1 (i.e., units with a given set of observable characteristics **x** have to belong to both the treated and untreated groups).

One of the most frequently used approaches to estimate average treatment effects under the assumption of selection on observables is Propensity Score Matching (PSM) [38,39]. In general, the basic idea of matching approaches is to determine a group of untreated units (*control group*) with similar values of the observable characteristics in **x** compared to those of the treated units. Then, an estimate of the ATET can be obtained as the mean of the differences between the observed outcomes and the counterfactual values:

$$AT\hat{\triangle}T\_M = \frac{1}{N\_1} \sum\_{i \in D\_i = 1} (Y\_{1i} - \hat{Y}\_{0i}) = \frac{1}{N\_1} \sum\_{i=1}^N D\_i \left(Y\_{1i} - \hat{Y}\_{0i}\right) \tag{5}$$

where *D<sup>i</sup>* = 1 identifies the set of treated units and the counterfactual outcome *Y*ˆ 0*i* is equal to *Y<sup>i</sup>* if *D<sup>i</sup>* = 0 and to a weighted average of the observed outcomes for the untreated units *j* chosen as matches for the treated unit *i*, ∑*j*|*D*=<sup>0</sup> *wijY<sup>j</sup>* , if *D<sup>i</sup>* = 0.

Rosenbaum and Rubin [37] suggested to match units according to the propensity score *π*(**x***i*) = *P*(*D* = 1|**x***i*), which is the conditional probability of receiving the treatment given the confounding variables **x**. In fact, if the CIA holds, it follows that (*Y*1;*Y*0)⊥*D*|*π*(**x**) and average causal effects can be thus estimated by conditioning on the propensity score *π*(**x**) instead of **x** (*Unconfoundedness Property*), reducing the multidimensionality of **x** to a single scalar dimension. The propensity score also entails that *D*⊥**x**|*π*(**x**), which implies that, conditionally on *π*(**x**), the treatment *D* and the observables covariates **x** are independent (*Balancing Property*). This property states that if *π*(**x**) is correctly specified, then units stratified according to the propensity score should be indistinguishable in terms of their observable characteristics **x**. Testing empirically the balancing property thus allows to assess whether the correct propensity score is being used.

The typical PSM procedure to compute ATEs consists of the following steps [40]:


In our empirical analysis, we use a probit specification to model a firm's probability to recur to alternative types of external support, as a function of the independent variables presented in Section 3.4, and estimate the propensity score. Then, following Heckman et al. [41], we use an Epanechnikov Kernel matching algorithm with automatic bandwidth selection [42], which matches every treated unit with a weighted average of all control units with weights that are inversely proportional to the distance between treated and control units. After testing for balancing, we estimate the ATET using Equation (5) to assess the impact of external support on a firm's adoption of resource efficiency measures.

#### **5. Results and Discussion**

#### *5.1. The Effect of External Support on Firms' Implementation of Resource Efficiency Practices*

To calculate propensity scores, we estimate alternative probit models to evaluate a firm's probability to recur to different types of external support, as a function of the observable characteristics discussed in Section 3.4. Probit estimation results are reported in Table 3. Estimated coefficients show that firm size has a positive and statistically significant effect on the recourse to external support, while firm age has a positive and significant impact only for the recourse to external advice, but it is not statistically significant for the other types of support. Firms with a turnover lower than 2 million Euro have a lower probability to recur to external support, whereas the availability of own internal financial resources and technical competencies tend to foster the combined use of external funding and advice, rather than single forms of support. Furthermore, we find that the firm's market segment significantly affects its propensity to recur to the different types of external support. Finally, we provide evidence of significant heterogeneity in the recourse to external support both across countries and over time.


**Table 3.** Probit estimates of the propensity to recur to external support.

**Notes**: The table reports results obtained from probit estimation of the conditional treatment probability on all the covariates for the different types of external support measures considered. Bootstrapped (200 replications) standard errors are reported in parentheses. \*\*\*, \*\* and \* denote significance of the parameters at the 1, 5 and 10% levels, respectively. **Source**: Own elaboration on Eurobarometer data.

Based on these probit estimations, we estimate propensity scores for each unit in the sample and we use an Epanechnikov Kernel matching algorithm, with automatic bandwidth selection and imposing common support, to match treated and untreated units and estimate average treatment effects. Table 4 presents the main results of the propensity score matching for the effect of the different types of external support on a firm's resource efficiency actions and investment behavior and on the benefits from the adoption of such actions (in terms of reduction in production costs) for the whole sample.


**Table 4.** Impact of external support on resource efficiency actions: ATETs for the whole sample of firms.

**Notes**: The table reports the ATETs of different types of external support measures on alternative indicators of resource efficiency actions and on the benefit from the implementation of RE actions (in terms of reductions in production costs), estimated on the whole sample of firms. The ATETs are computed using Epanechnikov kernel matching, with automatic bandwidth selection and imposing common support, performed using the Stata module *kmatch* by Jann [42]. Bootstrapped (200 replications) standard errors are reported in parentheses. Pseudo-*R* 2 is obtained from probit estimation of the conditional treatment probability on all the covariates on the matched sample. Mean bias and median bias are summary indicators of the distribution of the absolute standardized percentage bias for each covariate after matching. B and R are the standardized difference in the means and the ratio of the variances of the propensity scores between treated and untreated firms after matching, respectively. \*\*\*, \*\* and \* denote significance of the ATETs at the 1, 5 and 10% levels, respectively. **Source**: Own elaboration on Eurobarometer data.

Before discussing the estimated treatment effects, we focus on assessing the quality of the matching. The indicators reported in the bottom part of Table 4 provide support to the effectiveness of the matching procedure in restoring balancing in the covariates. In particular, the values of pseudo-*R* <sup>2</sup> decrease substantially after matching and almost approach zero for all the models. Similarly, the mean and median values of the absolute standardized percentage bias are remarkably lower after matching. Additionally, the values of both the standardized difference in the means (Rubin's *B*) and the ratio of the variances (Rubin's *R*) of the propensity scores between treated and untreated firms after matching are always within the ranges suggested by Rubin [43] for a balanced distribution of covariates (i.e., less than 25 for *B* and between 0.5 and 2 for *R*), providing support to post-matching balancing. Finally, Figure 1 reports kernel density plots for the estimated propensity scores of treated and untreated units before and after matching. For the matched sample, the plots are almost indistinguishable, confirming that matching on the estimated propensity score has balanced the covariates. Overall, the evidence obtained suggests that there are no systematic differences in the distribution of observable characteristics between treated and untreated firms after matching. ues of both the standardized difference in the means (Rubin's ances (Rubin's

**Figure 1.** Propensity score distribution for treated and untreated groups before and after matching. **Source**: Own elaboration on Eurobarometer data.

Once we have ensured that the balancing property is satisfied, we estimate the average treatment effects on the treated (ATETs), measuring the additional effect of external support on the firm's engagement in greening processes (in terms of adoption and investment in resource efficiency practices) and on the benefits from the adoption of these processes (in terms of production cost reduction).

Coherently with the findings of Bodas-Freitas and Corrocher [8] and in line with our first hypothesis, empirical results show that the overall effect of external support is positive and statistically significant both on the adoption and on the benefits from the adoption of

green practices. The additional effect of external support observed among treated firms is always statistically significant at the 1 and 5% level, with the only exception of the effect on the residual category *Other RE actions* and on the *Recycling Materials* process, which is significantly affected only by the recourse to external advice and by the combined use of both direct and indirect external support. The estimated ATETs show that firms recurring to any type of external support adopt, on average, a number of resource efficiency actions 0.426 higher than those in the control group. This value further rises to 0.65 when we consider the combined use of external funding and advice. The effect of external support is higher for the adoption of re-engineering practices (0.262 and 0.653, respectively, for any external support and for the combined use of external funding and advice) than for the adoption of waste management practices (0.164 and 0.233). Moreover, the additional impact of indirect support in the form of advice and consulting on the number of practices implemented by the firm is always slightly higher than that of direct financial support. Focusing on the ATET of external support on the different resource efficiency practices, we notice that firms relying on any form of external support are about 7% more likely to implement actions to save water and energy (the ATETs are equal to 0.067 and 0.073, respectively), to reduce waste (0.069) and to sell scrap materials (0.071); these effects further increase to about 9–10% when we consider the combined use of external financing and advice. It is also worth remarking that for these four resource efficiency practices, indirect support has a significantly higher additional effect than direct financing support. Interestingly, the use of renewable energies emerges as the only practice for the implementation of which external financial support exerts a much larger effect than external consulting: firms relying on direct (public or private) financing are, on average, 8.4% more likely to use renewable energies than firms that do not use this type of external support.

As in Hoogendorm et al. [14], we also assess the impact of external support on firms' investment in resource efficiency practices. The estimated ATETs point out that, on average, firms that rely on any type of external support, are 10.8% more likely to invest at least 1% of their annual turnover in resource efficiency activities than firms in the control group. The estimated treatment effect rises to 13.3% for the recourse to direct financial support, pointing out the key role played by access to external support in boosting a firm's investment in greening processes.

With respect to the impact of external support on firms' performance, in line with the previous literature [7,8,26], firms using external support are from 6.0 to 7.6% more likely to experience cost savings in comparison to firms with similar characteristics that do not rely on external support. This evidence confirms our second hypothesis, providing strong empirical support to the significant role of both direct and indirect support not only in fostering firm adoption of resource efficiency actions and investment, but also in increasing the cost-related benefits of implementing resource efficiency practices.

#### *5.2. Sensitivity Analysis: Firm Size and the Role of External Support on Greening Processes*

In this section, we explore the presence of heterogeneous effects of external support on firms' engagement in greening processes according to firm size. Tables 5 and 6 show the results of the propensity score matching for the subsamples of SMEs and large firms, respectively. Empirical results fully confirm our third research hypothesis, according to which the role of alternative forms of external support on the implementation of different types of environmental practices significantly differs according to firm size.


**Table 5.** Impact of external support on resource efficiency actions: ATETs for the subsample of SMEs.

**Notes**: The table reports the ATETs of different types of external support measures on alternative indicators of resource efficiency actions and on the benefit from the implementation of RE actions (in terms of reductions in production costs), estimated on the subsamples of SMEs. The ATETs are computed using Epanechnikov kernel matching, with automatic bandwidth selection and imposing common support, performed using the Stata module *kmatch* by Jann [42]. Bootstrapped (200 replications) standard errors are reported in parentheses. Pseudo-*R* 2 is obtained from probit estimation of the conditional treatment probability on all the covariates on the matched sample. Mean bias and median bias are summary indicators of the distribution of the absolute standardized percentage bias for each covariate after matching. B and R are the standardized difference in the means and the ratio of the variances of the propensity scores between treated and untreated firms after matching, respectively. \*\*\*, \*\* and \* denote significance of the ATETs at the 1, 5 and 10% levels, respectively. **Source**: Own elaboration on Eurobarometer data.


**Table 6.** Impact of external support on resource efficiency actions: ATETs for the subsample of large firms.

**Notes**: The table reports the ATETs of different types of external support measures on alternative indicators of resource efficiency actions and on the benefit from the implementation of RE actions (in terms of reductions in production costs), estimated on the subsamples of large firms. The ATETs are computed using Epanechnikov kernel matching, with automatic bandwidth selection and imposing common support, performed using the Stata module *kmatch* by Jann [42]. Bootstrapped (200 replications) standard errors are reported in parentheses. Pseudo-*R* 2 is obtained from probit estimation of the conditional treatment probability on all the covariates on the matched sample. Mean bias and median bias are summary indicators of the distribution of the absolute standardized percentage bias for each covariate after matching. B and R are the standardized difference in the means and the ratio of the variances of the propensity scores between treated and untreated firms after matching, respectively. \*\*\*, \*\* and \* denote significance of the ATETs at the 1, 5 and 10% levels, respectively. **Source**: Own elaboration on Eurobarometer data.

In the subsample of SMEs, the estimated ATETs remain positive and statistically significant (at the 1 and 5% level), confirming the estimation results obtained in the whole sample. For small and medium enterprises, coherently with Bodas-Freitas and Corrocher [8], direct financial support and indirect support in the form of advice and consulting significantly contribute to increase the number of green practices adopted, with estimated ATETs varying from 0.496 to 0.702. This result suggests that the two types of external support allow firms to overcome both the financial barriers hampering the acquisition of new equipment and technologies and the knowledge barriers related to lack of competencies and technical expertise necessary to implement resource efficiency practices. In particular, the combined use of external funding and advice exerts the largest effect on the number of re-engineering practices (0.454), while the external consultancy seems particularly important for waste management actions (0.257) and for the cost-related benefits from the adoption of resource efficiency measures. In this latter respect, those firms using external advice and consultancy are 8.7% more likely to experience a reduction in production costs compared to firms in the control group. Both types of external support also have a significant and positive impact on firm investment in greening processes: SMEs relying on direct and indirect support are about 11.8 to 16.6% more likely to invest at least 1% of their turnover to implement resource efficiency measures than similar SMEs not recurring to external support. As in the whole sample, external financing significantly contributes to increase the probability of resource efficiency investments of SMEs. Firms in the treated group have, on average, a 16.6% higher probability to invest in resource efficiency activities than firms in the control group. This probability increase is equal to 11.8% for those firms that receive external consulting and to 14.7% for those relying on both external financial and consulting.

Focusing on the subsample of large firms, most of the estimated ATETs lose their statistical significance, demonstrating that the additional effects of external support on the implementation of resource efficiency practices and on the benefits from the adoption are not particularly relevant. This evidence suggests that large firms tend to rely more on their internal resources for the adoption of green practices, having more financial resources to invest and better internal competencies and expertise. Specifically, for the subsample of large enterprises, external advice and consultancy is the type of support that contributes the most to the adoption of green practices, while direct financial support has an irrelevant impact on firms' engagement in greening processes. The impact of external advice strongly contributes to determine the significant effect of the combined use of direct and indirect support on the number of resource efficiency practices implemented by the firm. Large companies relying on both types of external support implement, on average, a number of resource efficiency measures 0.627 higher than those in the control group. As in the whole sample, the ATET is higher for the adoption of re-engineering practices (0.429) than for the adoption of waste management practices (0.208). It is also worth remarking that the combined use of external funding and advice significantly affects the cost-related benefits of adopting resource efficiency measures. Specifically, the estimated ATET indicates that larger firms relying on both direct and indirect support are, on average, 15% more likely to benefit from a reduction in production costs compared to larger enterprises in the control group.

#### **6. Conclusions**

This paper explores the impact of different types of external support on a firm's adoption of resource efficiency actions and investment behavior and on the benefits from the adoption of such actions (in terms of reduction in production costs). We rely on crosssectional data from the third and fourth waves of the Flash Eurobarometer survey "SMEs, resource efficiency and green markets", focusing on 6595 SMEs and large manufacturing firms from 35 European countries.

We use a propensity score matching approach, which compares the effect of the recourse to external support in the subsamples of treated and untreated firms with similar observable characteristics. In particular, we focus on the additional effect of external support on the adoption of resource efficiency actions and estimate the average treatment effect on the treated (ATET).

Our main results show that the overall effect of external support is positive and statistically significant, on both the adoption and the benefits from the adoption of green practices. Firms recurring to any type of external support implement, on average, a number of resource efficiency actions higher than those in the control group; the effect is particularly high for those firms that have jointly used external funding and advice. Even the probability to invest in resource efficiency activities is higher for those firms that can rely on external support, especially in the case of recourse to external financial support, demonstrating the relevance of external funding in boosting a firm's investments in greening processes. Moreover, the recourse to any form of external support fosters production cost savings, in particular, the simultaneous use of external financing and consultancy exerts a higher effect on the cost-related benefits.

The analysis on the subsamples of SMEs and large firms reveals significant heterogeneity in the impact of external financing and advice on environmental behavior. While the estimated ATETs for SMEs substantially confirm the results obtained in the whole sample, empirical results for large firms highlight that the additional effects of external support on the implementation of resource efficiency practices and on the benefits from the adoption are not particularly relevant. Large firms tend to rely more on their internal financial resources and expertise for the adoption of green practices. In particular, direct financial support has an irrelevant impact on firms' engagement in greening processes, while in the subsample of SMEs, it exerts a crucial role especially in boosting resource efficiency investments.

Our findings suggest that public policies, aimed at enhancing firms' involvement in greening processes, should be designed by taking into account firm size and different types of environmental practices. Specifically, for small and medium-sized firms, public and private financial support can directly improve the extent of resource efficiency investments, while external advice can contribute to integrate the lack of specific expertise and overcome the erroneous perception of environmental practices as an additional burden. On the other hand, the recourse to external advice and consulting services plays a particularly important role for large firms, enhancing the implementation of technological and managerial solutions that may improve the efficiency of environmental actions and encourage eco-innovation activities.

Our analysis has some limitations, mainly related to the data used in the empirical analysis, that need to be acknowledged. First, the cross-sectional nature of the data does not allow to fully control for unobservable heterogeneity at the firm level. At the same time, it prevents any attempt to investigate the intertemporal relationship between firms' implementation of resource efficiency practices and the recourse to direct and indirect support. Future analyses should exploit longitudinal firm-level data to estimate the average treatment effects of external support on firms' environmental practices and assess the validity of the evidence obtained in this study. The use of panel data will also allow to develop an intertemporal framework to analyze firms' environmental behavior. Furthermore, as pointed out by Hoogendoorn et al. [14], the available data do not allow to distinguish between stakeholder groups nor to identify the specific products or services offered by the firm. This prevents properly assessing stakeholders' influence and the effect of firm tangibility on the extent and types of environmental practices and on the related recourse to external support. Finally, our study provides first empirical evidence on the heterogenous effects of external support on the environmental practices of SMEs and large firms. Future research efforts are needed to shed additional light on these firm-size heterogeneities.

**Author Contributions:** D.A. and M.G. equally contributed to the development of this research. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was conducted within the project "Bank management, finance and sustainability" financed by the University of Perugia (Fondo Ricerca di Base, 2019).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A**

**Table A1.** Variable definitions.



#### **References**


### *Article* **The Impact of Mobile Money on the Financial Performance of the SMEs in Douala, Cameroon**

#### **Frank Sylvio Gahapa Talom and Robertson Khan Tengeh \***

Department of Entrepreneurship and Business Management, Faculty of Business and Management Sciences, Cape Peninsula University of Technology, Cape Town 7535, South Africa; 212021850@mycput.ac.za **\*** Correspondence: tengehr@cput.ac.za; Tel.: +27-21-460-3450; Fax: +27-86-778-0394

Received: 25 November 2019; Accepted: 22 December 2019; Published: 24 December 2019 -

**Abstract:** Often financially excluded by the traditional banking system, small and medium-sized enterprises (SMEs) in many developing countries have found in mobile money services (MMS) a sustainable alternative. Despite its potential in propelling inclusive growth, the use and adoption of mobile money (MM) by SMEs has generally been low in developing countries, and one of the reasons has been limited data that supported its impact on financial performance. As a result, there was a need to investigate the impact of the mobile money payment and receipt services on the financial performance of the SMEs in Cameroon. This paper implemented a mixed research paradigm with data collected through the administration of a survey questionnaire and from one-on-one in-depth interviews. A sample of 285 SMEs responded to the survey, while 12 owners/managing directors were purposively selected to participate in the personal interviews. Version 25 of the Statistical Package for the Social Sciences (SPSS) software was used to analyse the quantitative data, while the qualitative data was analysed along themes. The results were, after that, triangulated for credibility reasons. The concluding findings indicated that the mobile money payment and receipt services contributed of the order of 73% of the total variance in the turnover of the SMEs in Douala after they had begun to use the technology. By confirming the positive relationship between the use of mobile money services and the financial performance of businesses, it is hoped that all the relevant stakeholders will see this as a possible solution to the financial challenges that SMEs face in developing economies.

**Keywords:** mobile money; SMEs; financial performance; payments and receipts; Douala, Cameroon

### **1. Introduction and Background**

Small and medium-sized enterprises (SMEs) make significant contributions to driving the economies of a great many countries. They play a crucial role in socio-economic development by contributing to the creation of wealth, economic growth and employment [1,2]. A 2016 census of enterprises in Cameroon suggested that approximately 99.8% of the enterprises in the country are SMEs [3]. Furthermore, it was revealed that SMEs accounted for 72% of the permanent jobs generated in Cameroon [3]. Although they employ around 72% of the workforce in Cameroon and contribute approximately 35% of the GDP [4,5], very few are considered to be structurally and financially stable [6].

The SME sector of Cameroon comprises mainly (around 97%) sole proprietorship or family businesses [3]. It was further noted that approximately 85% of the managers of these SMEs were relatively poorly educated. Because of their structure and the managerial profile of their managers, SMEs tend to be highly cash-dependent [7–10]. Consequently, they are usually obliged to transact business with suppliers, to buy or pay for goods by travelling to their offices, which can entail considerable risk of theft or losing money. In the case of SMEs which have bank accounts, apart from the disadvantages of costs which are incurred by high bank charges, documentation and transport, owners are frequently required to queue for lengthy periods before they can obtain access to funds, which makes it very difficult to exploit any unexpected opportunities which may arise for which funds are required [11]. Because the owners of many of the businesses are sole traders who operate in a very informal manner and the businesses are often staffed only by their owners and possibly one or two members of their families [12,13], they are often obliged to leave their businesses unattended for several hours in order to conduct transactions in a bank [11]. As a consequence, sales are lost, and their prospects for survival are severely compromised [14,15].

As it has become abundantly evident that to survive and achieve growth, SMEs in Cameroon need to streamline their procedures, reduce operating cost and eliminate unnecessary loss of time, the advent of the phenomenon of mobile money could not have been more fortuitous [16]. The system enables the SMEs to receive payments directly from customers and also to make payments directly to suppliers through mobile telephones [17], without being obliged to leave or close their premises for lengthy periods. Accordingly, it provides a viable means for people or SMEs who do not have access to bank accounts to make financial transactions with ease, have access to funds when they are needed, without incurring additional charges such as transport and opportunity cost, and significantly improve the performance of their businesses as a consequence [18,19].

According to Ngaruiya et al. [10], obstacles are inherent in the operations of SMEs. In the case of Cameroon, SMEs needs concerning financial liquidity and banking services are not sufficiently met by commercial banks for several reasons, including a lack of collateral, inadequate bookkeeping systems, and their often questionable viability in the eyes of financial institutions [20–22]. Besides, SMEs bank accounts are not cost-effective, owing to high bank charges and the transport costs which are incurred by travelling to banks to make transactions [10,23–25]. These unwieldy procedures have contributed to the performance of many SMEs in Cameroon stagnating, with low economic growth being but one of a host of adverse consequences. As SMEs comprise the majority of the businesses in the country and in the light of the disastrous effects which cumbersome banking procedures have on their performance, a strong case could be made for the use of mobile money as a means of enabling SMEs to streamline their operations. The platform will improve the mode of receipts and payments, the debt collection procedures which, in return, will enhance the liquidity and working capital management problems faced by the SMEs [23].

Although mobile money does not provide a panacea for all of the financial problems with which SMEs are faced, the benefits far exceed the disadvantages which are associated with adopting the system. Irrespective of whether the system is used in isolation or conjunction with a bank account, it stands to increase the sales of SMEs and reduce their operating costs, with both factors making positive contributions to improving their financial performance [10].

Although a considerable amount of research has been conducted concerning the effects which mobile money has had upon the performance of SMEs in Africa, particularly in Kenya, the topic has not been examined in Douala, Cameroon. The studies that delved in this area include Ngange and Beng [19] that studied the impact of mobile phone usage in economic development in Molyko. Covering a much bigger area, Ojong [26] looked at informal mobile remittances and socioeconomic factors in the North-West Region. Mwafise and Stapleton [27] examined the influence of social–technical and institutional factors on the effective uptake of mobile money electronic payments. Yet, none of the just-mentioned studies investigated the impact of the mobile money services on the performance of SMEs in Douala.

More so, this paper focused principally on the payment and receipt services. Consequently, this study aims to contribute to the mobile money literature on Cameroon by mainly assessing the impact of the mobile money payment and receipt services on the financial performance of the SMEs in Douala, Cameroon. The paper adopted the mixed research method to record and analyse quantitative and qualitative data. The individual methods were further triangulated to increase the credibility of results. In line with similar studies, such as Ngaruiya et al. [10], Mararo and Ngahu [11], Masocha and Dzomonda [18] and Higgins et al. [23], it is hoped that the findings of this paper will be useful to the relevant stakeholders in Douala, Cameroon.

The rest of the paper is structured as follows. A review of the literature related to mobile money, SMEs and the possible opportunities and challenges associated with the platform. After that, a discussion around the research design methodology, followed by discussions of findings. Finally, the conclusions, limitations of the study and recommendations were provided to the relevant stakeholders. Because of the boundaries of the study, suggestions for future research directions were pointed out.

#### **2. Literature Review**

#### *2.1. Mobile Money and SMEs*

Mobile money (MM) is a service which permits customers to obtain access to financial services employing cellular devices [28], by dialling Unstructured Supplementary Service Data (USSD) codes. USSD is a communications protocol for mobile communication technology which is used to send text between mobile telephones and an application programme in the mobile network, which does not require users to have access to the internet. Although the technological innovation is now available in many developing countries, its use is particularly widespread in countries in which it is difficult for many citizens to open bank accounts and/or access banking services [22,29]. It enables users to store, send and receive money without the transactions entailing the use of bank accounts [30].

Mobile money has disrupted the financial sector and the way of transacting. SMEs can now efficiently conduct financial transactions, anytime, and anywhere, without necessarily having a bank account [10,24,26,30]. This innovation can help to reach those who do not have access to banking services and thus improve financial inclusion [22]. Ngange and Beng [19] and Chimaobi and Chizoba [31] demonstrated that mobile commerce facilitates communication between users. The later scholars went further to show that the platform improves the efficiency of the business operations. Additional studies conducted respectively by Ngaruiya et al. [10], Higgins et al. [23] and Mbogo [25] proved that mobile money improves business networking, while Amponsah [21], Chimaobi and Chizoba [31] and Ngaruiya et al. [10] demonstrated that the technological innovation promotes a cashless economy. All these benefits taken concurrently will enhance the productivity, decrease the operating costs and thus improve the performance of the SMEs [10,18,32,33].

In the Economic and Monetary Community of Central Africa (CEMAC) region, which is heavily regulated, mobile money is still at an embryonic stage [29]. Mobile money services providers are bound to work with their partner banks to provide their services [30]. This seems to be common practice in many emerging economies, with some exceptions, like in Ghana, where, since 2015, the mobile network operators (MNOs) can now apply for licenses directly from the Central Bank [21]. Mobile network operators are telecommunication entities that provide services for mobile phone subscribers. Also, the mobile money services which are provided are limited by comparison with those which are available in East and West African countries [10,11,33]. In Cameroon, for instance, mobile money is used mainly to make purchases and send and receive money; saving and loan facilities are not yet provided [34,35].

The study identifies four significant platforms for mobile money services (MMS) in Cameroon, namely, MTN Mobile Money, Orange Mobile Money, Express Union Mobile Money and the recently launched Nexttel Possa. By contrast, there are only two service providers, namely, MTN and Orange Cameroon, also known as mobile network operators, which dominate the Cameroonian mobile money market and together account for 5.4 million registered users [36]. Orange, which has 2.8 million registered users, offers the following services: money deposits, money withdrawals, the sending and receiving of money transfers, a Visa card facility, the purchasing of insurance-related products, the transferring of funds between bank accounts and mobile money accounts, the purchasing of airtime and the payment of bills, university fees, transport tickets and school fees [35]. MTN, which has 2.6 million users, offers a similar range of services, with the exception of the Visa card facility, the purchasing of insurance-related products and the transferring of funds between bank accounts and

mobile money accounts [34]. According to FinMark Trust [22], the Cameroonian population comprised more than 14 million people who were 15 years of age or older in 2017. Combined with the figure of 12% citizens who hold bank accounts [29] and the average population growth of 2.7% in Cameroon between 2014 and 2017 [37], it appears that in the region of 1.7 million people held bank accounts, which amounts to around one-third of the 5.4 million registered users of mobile money during the same period. As mobile money is a relatively new phenomenon in Cameroon, it is highly significant that three times more Cameroonians have opened mobile money accounts than hold bank accounts. Similar findings were obtained in at least eight countries [38].

According to Rubini [39], SMEs are considered to be the backbone of most of the developed and developing countries across the globe. The term "small and medium-sized enterprises" is a broad one, and the specific defining attributes tend to vary among individual countries. The categorisation of enterprises with respect to their size on the basis of the numbers of workers which they employ, their annual turnover or capital assets entails fairly arbitrary assessments, which are often influenced by the prevailing business values of individual countries [10]. To cite an internationally accepted criterion, the Organisation for Economic Cooperation and Development classifies SMEs as businesses which do not employ more than 249 employees [1]. For practical purposes, having the number of employees as a defining criterion provides a useful comparative measure for assessing the sizes of businesses [1]. In Cameroon, the official definition is derived from the law No 2010/001 of 13 April, 2010 for the promotion of small- and medium-sized enterprises [40]. It holds that any company with an annual turnover (excluding tax) which does not exceed FCFA 1 billion and employs a permanent workforce of not more than 100 employees is considered to be an SME [41]. The FCFA is denomination of the common currency of 14 African countries which are members of the Franc Zone. As per their contribution into the economic and the social well-being across the globe [1,39], Cameroonian SMEs play a crucial role in achieving economic growth by contributing up to 72% of the national workforce [3] and approximately 35% of the GDP [4]. Despite their important responsibility, the majority in emerging countries face many difficulties, with the most important being that of financing [3,4,14,42]. In developing countries such as Ghana, Tanzania and Kenya, credit systems have been developed by MNOs which are based upon transactional histories of mobile money, which make it possible to grant microloans to SMEs [43]. In Cameroon, it is the State that made an effort by opening a bank for the SMEs in July 2015 [44]. However, even if the MNOs cannot offer microcredit in Cameroon, it is perceived by many scholars that the mobile money services facilitate the commercial dealings for SMEs [19,31]. According to Ngaruiya et al. [10] and Amponsah [21], the rapid diffusion of mobile money transfer is seen as a potentially vital tool for facilitating financial transactions. This indicates that the rapid adoption of mobile money Services is seen as a way to improve the financial functionality and hence the performance of the SMEs. For Pinem and Dwi [45], the performance of the SMEs can be measured by evaluating the sales growth, which remains one of the main determinants of SMEs performance [46]. The significance of the MMS seems to be mitigated in a good number of countries due to delays in telecommunication infrastructures [47]. Nevertheless, as demonstrated by Ngaruiya et al. [10], Masocha and Dzomonda [18], Higgins et al. [23], Mbogo [25], Chale and Mbamba [32] and Nyaga and Okonga [33]—in their respective studies, the platform has improved the financial performance of SMEs after they have begun to use it.

#### *2.2. Opportunities*

Apart from the role which the system has played in increasing rates of financial inclusion, businesses which adopt mobile money services benefit from a wide range of different advantages and opportunities [21]. Among the many other advantages which it provides is the ability to transfer money at a low cost within a branchless bank [10,33]. Evidence from studies carried out in Kenya and Ghana attest to how the use of MM ensures a seamless cash flow, easier and safer financial transaction for SMEs [10,21]. The concept of a cashless economy is enthusiastically promoted by many central banks throughout Africa [10,21,31]. Ngaruiya et al. [10] point out that the adoption of the system

has facilitated decision-making and the exchange of information, improved the ability of businesses to network successfully and increased the competitiveness of SMEs. The findings of a study which was conducted by Chimaobi and Chizoba [31] revealed that SMEs in Nigeria, which traded using mobile systems, were able to shorten their delivery times significantly. Both Ngange and Beng [19] and Chimaobi and Chizoba [31]—maintain that using mobile money services facilitates communication between users and improves relationships between buyers and sellers. Effective communication has also significantly reduced the effects of the phenomenon which is known as asymmetric information or information failure between users, which results from one user in a transaction having significantly more information on it than the other. Increasing the range of opportunities which are available to users via the platform would create added value for SMEs and enable them to reduce their operational expenses and, indirectly, improve their performance and growth [18,23,25,32,33]. Despite the immense opportunities which mobile money services provide to users, groups of factors continue to militate against their universal adoption.

#### *2.3. Di*ffi*culties*

The principal categories of factors which tend to discourage the universal adoption of mobile money services aremainly regulatory, infrastructural and those which arise from traditional perceptions [24,39,48–51]. Mobile money remains heavily regulated [24,39]. In some instances, stakeholders in the traditional banking sector tend to perceive the new system as a threat to the hegemony which they have maintained [24] and do not welcome the prospect of their services being superseded or supplanted by innovative contemporary ones [21]. The pressure which commercial banks are placing upon the central banks of their countries leave them with two possible courses of action: continue with the status quo and retard economic growth as a consequence of stifling the growth of SME sectors or liberalise and permit newcomers to the financial sectors of their countries to boost national economies by providing services which significantly increase the financial performance of SMEs [21,24,39,48]. Limited infrastructure has made it impossible to make mobile money services available to all of the members of the populations of some countries [42,47]. As has already been noted, cellular coverage tends to be low in many developing countries [49]. As it is particularly low in rural areas, many people are effectively denied access to the advantages which cellular technology provides [49]. Unstable networks and interrupted transmission oblige some users of mobile money services to travel to locations in which their networks are functioning normally to make transactions, thereby incurring additional costs and suffering considerable inconvenience [47]. According to Chimaobi and Chizoba [31], the erratic transmissions of energy by the power supplier also affect users of mobile money services adversely. During blackouts, it is impossible to make transactions (mobile signals blackout), and, in some instances, cellular devices are damaged beyond repair (electricity blackout). Significant numbers of members of the populations of developing countries tend not to trust modern technology and prefer to carry cash with them, owing to the degree of control over their transactions which they perceive that doing so provides [47]. For instance, it is not possible to request a refund or to stop a transaction which has already been validated in a mobile money transaction, while it is easy to do so in the case of cash payment in a supermarket. Moreover, in Pakistan, no matter the level of education, people still prefer to keep money at home [51].

#### *2.4. Mobile Money Evolution in Cameroon*

Mobile money was first launched in Cameroon in 2011. The Cameroonian subsidiaries of telecommunication leaders MTN and Orange pioneered the concept and officially launched it in 2012 [52]. The circumstances which prompted its launching were similar to those of most developing countries, particularly concerning the small numbers of members of the population who held bank accounts [29]. As had been the case in the other countries in which the concept had been launched, many households and SMEs in Cameroon had been effectively excluded from the traditional banking system and without access to funding in the formal sector [3,4,42].

Although the services which mobile money provides in Cameroon do not include financing now, its introduction had significantly increased the financial inclusion rate (29%) by 2017 [22], from 9%

in 2012 [53]. As a direct consequence, many citizens have been able to ply trades and launch startup enterprises, which have resulted in indirect employment for of the order of 5000 people [54]. The mobile money transactions which have accompanied this surge amount to in the region of FCFA 3500 billion in 2017, a figure which represents 17.5% of the GDP of Cameroon [53]. This represents an increase of more than 1000% from the FCFA 300 billion recorded in 2016 [53].

The introduction of mobile money has enabled Cameroonian households to incur reduced costs by saving and reduce the risk of loss and theft which had accompanied saving in the past [21,22,24,48]. As the mobile telephone penetration rate was 71% in 2014 [52], and that of holding bank accounts had been one of the lowest in the world at 12% [29], it is abundantly evident that mobile money could not have arrived in Cameroon at a more promising time. The mobile money service in Cameroon is provided through a partnership between commercial bank and mobile network operators (MTN Cameroon, Orange Cameroon, CAMTEL, and Nexttel) because only commercial banks are allowed to issue electronic money [30], and the mobile network operators own the telecommunication infrastructures and technologies to deploy the platform. This regulating arrangement of convenience is the status currently prevailing in Cameroon and will surely deter the significance of the MM in the long run.

Although its importance is affected by factors related to regulation, infrastructures and customs, Mobile Money appears to be the solution to the multiple problems, namely, liquidity, means of payments, debt collection, working capital and financing faced by SMEs. Its adoption and usage in their day-to-day activities have had a positive impact on their performance, as shown by many scholars.

#### **3. Research Design and Methodology**

This study opted for the pragmatism paradigm. The positivism and interpretivism philosophies were adopted in order to collect data from SMEs in "Mboppi" and "Central" markets. The researcher elected to make use of both quantitative and qualitative research methods in this study to obtain as complete an understanding as possible of the research problem and to make effective use of any converging information which the quantitative and qualitative studies generated. The strategy also enabled the researcher to perform a rigorous evaluation of the reliability of the findings by using the qualitative findings to corroborate the results which the survey questionnaire generated through triangulation.

#### *3.1. Sampling Technique*

In this study, the researcher was unable to determine, with an acceptable degree of exactitude, how many of the SMEs which qualified for selection in the two markets were using mobile money services to make and receive payments. Also, the criteria concerning the periods for which the SMEs had been operating and the numbers of employees which they had made it even more difficult to identify and obtain access to potential participants. Using simple random sampling would have entailed an unacceptable degree of difficulty, been excessively time-consuming and entailed expense which the researcher could not afford.

Polit and Beck [55] explain that sampling is a method of choosing a portion of a target population to represent the population as a whole in the respects in which particular researchers are interested in the purposes of their studies. The researcher elected to use nonprobability sampling to select participants who were readily identifiable as fulfilling the criteria for inclusion in the research sample and drew upon their knowledge of local SMEs to locate other potential participants through snowball sampling [55,56].

Mindful that qualitative phase seeks to understand better the underlying reasons and motivations rather than to quantify and generalise to a broader population, it is inappropriate to use random sampling techniques [57]. The participants for the in-depth interviews were purposively selected from among the respondents to the questionnaire based on their sales turnover, to obtain a research sample whose members represented SMEs whose turnover ranged from the minimum to the maximum levels and also adequately served those with intermediary levels of turnover. The survey questionnaire was administrated over 12 weeks, between November 2018 and January 2019; while in-depth interviews took around six weeks, from February to March 2019.

#### *3.2. Sampling Size*

Conscious of many restrictions including time, finance and limited access and the fact that the population as a whole is too large to work with, the researcher was not able to collect or analyse data from the entire population. As Dudovskiy [58] maintains that a sample size of twelve is sufficiently large for a qualitative study of a homogeneous population, the researcher selected to conduct twelve in-depth interviews.

For the quantitative component, the researcher encountered a considerable amount of difficulty in determining an optimal sample size for the administration of the survey questionnaire, in the absence of official statistics concerning the numbers of the SMEs which were making use of mobile money services in the two markets, and even more difficulty in identifying SMEs which had been operating for two years or more. Consequently, the researcher elected to use the formula which Cochran [59] developed to calculate the size of the research sample for the quantitative study: *n*<sup>0</sup> = *Z* <sup>2</sup>*pq e* 2 , where n<sup>0</sup> is the sample size, Z<sup>2</sup> is the abcissa of the normal curve that cuts off an area α at the tails (1−α equals the desired confidence level, e.g., 95%), e is the margin of error, p is the projected percentage of a characteristic which is to be found in a population and q is 1−p. The value of Z is found in the statistical tables which contain the area under the normal curve [60].

Following the application and the computation of the Cochran's formula [59] at a confidence level of 95%, a margin error of 5% and a standard deviation of 50%, the researcher obtained a sample size of 384 for the quantitative phase of the study. The sample size of 384 was considerably greater than those who had been used in similar studies which have been conducted in countries in which the use of mobile money services is widespread. The researcher calculated an average sample size of 228 by consulting the literature pertaining to the studies which had been conducted by Ngaruiya et al. [10], Mararo and Ngahu [11], Higgins et al. [23] and Nyaga and Okonga [33]. After having given due consideration to the relatively new status of the mobile money industry in Cameroon, the relatively limited adoption of money market services in Cameroon by comparison with Kenya, the relatively short lifespans of many SMEs in Cameroon [6] and the constraints which time and financial considerations imposed, the researcher decided upon a sample size of 250 for the quantitative study. To compensate for any unusable questionnaires and to ensure that the final sample size was as close to 250 as possible, the researcher distributed a total of 300 questionnaires evenly among potential respondents in both markets. After he had collected and sorted the completed questionnaires, it emerged that 285 were usable, of which 142 had been completed by respondents in the Central Market and the remaining 143 by respondents in the Mboppi market. As the 285 completed questionnaires significantly exceeded the initial target figure of 250, it was likely that the credibility of the findings would be substantially increased.

#### *3.3. Measures Taken to Ensure the Credibility of the Findings*

Although it is not possible to eliminate the possibility of the findings of research studies lacking credibility, researchers need to take all reasonable measures to do so [58]. Credibility refers to the extent to which accounts which are provided by researchers are plausible and appropriate, particularly concerning the degree to which their findings accord with the perceptions of the participants in their studies [55]. Credibility is predicated upon the criteria of reliability and validity to evaluate the quality of research.

#### 3.3.1. Reliability

According to Dudovskiy [58] and Asoba [61], reliability refers to the consistency with which particular research instruments generate data. Consequently, the reliability of the findings of a study is assessed in accordance with the likelihood that other researchers would be able to generate similar findings under the same conditions and using the same research techniques. As such, the reliability of the findings of this paper was ensured by pilot testing both the survey questionnaire and the interview guide and by subsequently corroborating the findings of the quantitative study with those which were obtained from the face-to-face in-depth interviews.

#### 3.3.2. Validity

According to Polit and Beck [55], validity can be defined as the degree to which a research instrument measures what it is intended to measure. From a slightly different standpoint, Dudovskiy [58] evaluates the validity of findings as a measure of the degree to which the requirements of a particular scientific research methodology have been adhered to during the process of generating research findings. In both instances, it is evident that validity is a measure of accuracy. Creswell [62] explains that in mixed methods research, the findings from quantitative studies are used to validate those of qualitative studies and vice versa. As this study employed a mixed methods research design, the findings from the administration of the survey questionnaire were validated against those which the in-depth interviews generated.

#### **4. Findings and Discussions**

#### *4.1. Results of the Quantitative Phase*

#### 4.1.1. Monthly Levels of Turnover in FCFA Before and After the Adoption of Mobile Money Services

Figure 1 depicts the levels of monthly turnover, which the respondents claimed on behalf of their SMEs before and after they had elected to make use of mobile money services. The ranges into which their respective levels of turnover fell are summarised in Table 1. The results in Table 1 denote that the average turnover of the SMEs in Douala increased by 0.44, almost 12%, after they have begun using mobile money services. It can be seen in Figure 1 that before they elected to make use of mobile money services, 69 out of 285 SMEs had achieved monthly turnovers of from FCFA 200,001 to FCFA 500,000, the figures rose to 75 after they had done so. Also, the numbers of SMEs which achieved monthly turnovers of more than FCFA 1 million rose from forty, before the adoption of mobile money, to sixty, after having done so, which represents an increase of 50%. The findings of a study which was conducted by Ngaruiya et al. [10] in Kenya were essentially similar to those of this study in these respects.

**Figure 1.** Monthly turnover figures in FCFA before and after the adoption of mobile money services (MMS) (Source: authors).



#### 4.1.2. Perceptions of the Respondents of Mobile Money and Mobile Money Services

Figure 2 depicts the distributions of the responses of the respondents according to a 5-point Likert scale concerning their perceptions of mobile money and mobile money services. A significant majority (67%) of the respondents either agreed or strongly agreed that it was affordable and straightforward to register a mobile money account. A similar majority (65%) either agreed or strongly agreed that mobile money transactions were safe, while (54%) either agreed or strongly agreed that mobile money service providers were reliable. By contrast, the majority (130) either disagreed or strongly disagreed that using mobile money services to make and receive payments had significantly influenced the turnover of their businesses, while ninety-two respondents chose to record neutral responses, by comparison with a total of sixty-three who either agreed or strongly agreed with the statement. The perceptions of the respondents were significantly skewed towards either negative or neutral responses to the statement that their sales had increased after they had begun to make and receive payments by means of mobile money, as they accounted for 196 of 285 responses. The spread of responses to the statement that the adoption of mobile money services to make and receive payments had improved their cash flow was similar, although an even larger group of 107 recorded neutral responses. There was a significant consensus of 242 respondents who either agreed or strongly agreed that mobile money payments and receipts reduced transport costs, while 198 either agreed or strongly agreed that they reduced the cost of transactions. A further 133 either agreed or strongly agreed that they reduced opportunity costs, while 184 either agreed or strongly agreed that they were more cost-effective than the services of banks.

**Figure 2.** Perceptions of the respondents of Mobile Money (MM) and Mobile Money services (Source: authors).

**'**

explains, Cronbach's alpha

Pearson's

**'**

4.1.3. Evaluation of the Perceived Impact of Mobile Money Payments and Receipts on the Financial Performance of SMEs in Douala

The researcher identified nine variables from the data to form a reliable scale against which to assess the influence of Mobile Money and Mobile Money services upon the financial performance of the SMEs whose representatives responded to the survey questionnaire. The researcher evaluated their reliability by using the SPSS version 25 software to determine a Cronbach alpha score for each before commencing with the analysis to test the hypotheses in the study. The nine variables, along with their respective Cronbach alpha scores, are summarised in Appendix A, Table A1. As Goforth [63] explains, Cronbach's alpha α scores need to be at least from 0.65 to 0.8 if they are to denote significance. The alpha coefficient noted in Table 2 is 0.659, which suggests that the items exhibit a reasonable degree of internal consistency concerning reliability. Reliability refers to the consistency with which particular research instruments generate data and is assessed by the likelihood that other researchers would be able to generate similar findings under the same conditions and using the same research techniques [58,61].


After that, Pearson's correlation analysis was conducted. For the purposes of performing the analysis, the variables have been recoded as follows: YE for years for which the SMEs of the respondents had been in existence, BA for does your business hold a bank account? AY for number of years since the adoption of MMS, MMSCETB for MMS are more cost effective than the services of traditional banks, MMSIT for using Mobile Money services influences turnover, TBA for monthly turnover before the adoption of MMS, TAA for monthly turnover after the adoption of MMS, NOPDS for number of Mobile Money payments per day to suppliers, EDU for levels of educational attainment and BS for business sectors in which the SMEs of the respondents operated. As can be seen in Table 3, the values of the correlation coefficient (*r*) for correlations between the variables which were regressed ranged from 0.149 to 0.834. The values reveal that there were correlations between all variables and turnover, with the exceptions of BA, EDU, and BS. It can be drawn from Table 2 that there was a significant and robust correlation at the 1 percent level between monthly turnover before the adoption of Mobile Money services and afterwards. Although there were correlations with all of the other significant variables, they were relatively weak. It was meaningful to note that the strong positive relationship between the variables TBA and MMSCETB—which was reflected in the value of 0.191 at the 1 percent level, dropped sharply after the adoption of Mobile Money services. This finding suggests that although many of the respondents may have believed before they started to use Mobile Money to make and receive payments that the availability of a more cost-effective system would increase their turnover, their perceptions changed after they started to use the system. It is also significant that the strength of correlations increased after the adoption of Mobile Money services for the variables YE, MMSIT and NOPDS. It is possible to infer from them that as the turnover of the SMEs increased over time, their increasing numbers of payments to suppliers per day also increased their turnover, thereby contributing to perceptions of Mobile Money transactions increasing turnover. This scenario also emerged from the findings of earlier studies which were conducted by researchers such as Ngaruiya et al. [10], Higgins et al. [23] and Nyaga and Okonga [33] -. Once the correlations had been determined between the independent and dependent variables, the researcher elected to investigate the proportion of variance in the dependent variable accounted for by the independent variables using the regression analysis. The R-square value of 0.733 in the model summary in Table 4 suggests that taken together, the independent variables explained of the order of 73 percent of the total variance in the turnover of SMEs in Douala after the adoption of Mobile Money services. From this

finding, it would appear that the independent variables which have been cited collectively constitute a credible predictor of financial performance for SMEs in Douala.



\*\*. Correlation is significant at the 0.01 level (two-tailed). \*. Correlation is significant at the 0.05 level (two-tailed). Years in existence (YE), Does your business hold a bank account? (BA), Years since adoption of MMS (AY), MMS more cost-effective than services of traditional banks (MMSCETB), MMS influences turnover (MMSIT), Monthly turnover in FCFA before adoption of MMS (TBA), Monthly turnover in FCFA after adoption of MMS (TAA), Number of payments per day using Mobile Money (NOPDS), Levels of educational attainment (EDU), Business sectors in which the SMEs of the respondents operated (BS). Source: authors.

**Table 4.** Model summary**<sup>b</sup>** of regression analysis.


<sup>a</sup> Predictors: (Constant), Number of payments per day using Mobile Money (NOPDS), Years in existence (YE), MMS influences turnover (MMSIT), MMS more cost-effective than services of traditional banks (MMSCETB), Years since adoption of MMS (AY), Monthly turnover in FCFA before adoption of MMS (TBA). b. Dependent Variable: Monthly turnover in FCFA after the adoption of MMS (TAA). Source: authors.

From the regression findings which appear in Table 5 and after substitution of coefficients (β . . .) and variables (*Y and X* . . .) onto the generic regression equation (*Y* = β0 + β1*X*1 + β2*X*2 + β3*X*3 + β4*X*4 +β5*X*5 + β6*X*6), the researcher obtained the following function:

$$Y = 1.098 + 0.108 \,\text{X} \\ 1 - 0.123 \,\text{X} \\ 2 - 0.154 \,\text{X} \\ 3 + 0.152 \,\text{X} \\ 4 + 0.752 \,\text{X} \\ 5 + 0.243 \,\text{X} \\ 6 \\ \tag{1}$$

where Y is the dependent variable (turnover after adoption of Mobile Money), X1 the independent variable 1 (years of existence), X2 the independent variable 2 (years since adoption of MMS), X3 the independent variable 3 (Mobile Money services are more cost-effective than services of traditional banks), X4 the independent variable 4 (MMS influence turnover), X5 the independent variable 5 (monthly turnover before adoption of Mobile Money) and X6 the independent variable 6 (number of payments to suppliers per day using Mobile Money).



<sup>a</sup> Dependent variable: monthly turnover in FCFA after the adoption of MMS (TAA). Source: authors.

The findings which are summarised in Table 5 confirm those which appear in Table 3, in that all of the independent variables apart from MMSCETB correlate positively with the turnover after the adoption of Mobile Money. All of the independent variables, apart from years since adoption, are significant at the 5 percent level. It needs to be emphasised that the findings suggest that a unit increase in the monthly turnover of SMEs in Douala before the adoption of Mobile Money to make and receive payments should result in a 75 percent increase in their financial performance after the adoption of the system, ceteris paribus. They also reveal that unit increases in the numbers of payments to suppliers per day and the perception that Mobile Money services are more cost-effective than those of traditional banks result in an increased coefficient value of 0.243 and a decreased one of −0.154, respectively, for the independent variables if the coefficients for other variables are kept constant.

#### *4.2. Results of the Qualitative Phase*

"

"

Has the adoption of Mobile Money Services to make and receive payments improved your business operations?

The question of whether the adoption of Mobile Money to make and receive payments had improved the operations of the interviewees drew mixed responses as shown in Figure 3, with 58 percent perceiving an improvement while the remaining 42 percent did not. In the words of interviewee A: "

"There is no change. I am still making more or less around the same turnover. The only advantage I may acknowledge is that it makes it unnecessary for me to leave my business premises."

This response confirmed that convenience was a motivating factor for the adoption of Mobile Money to make and receive payments. As interviewee D explained,

"The majority of my customers use Mobile Money services, especially those outside of Douala. Mobile Money services have improved my operation and made it very fluid. Now, people can pay for goods from wherever they are in Cameroon and receive them." "

It needs to be emphasised that improved business operations are likely to result in increased sales turnover, a definitive indicator of both growth and financial performance. The overall finding displayed in Figure 3 was that the acknowledgement of seven of the twelve interviewees that the adoption of Mobile Money for payments and receipts had improved their business operations represented an acknowledgement that doing so had improved the financial performance of their businesses.

#### *4.3. Triangulation of Results from Both Phases*

The triangulation process helps to confirm the interpretations which researchers have made of their data [62,64]. The findings which were obtained from the personal interviews in the qualitative study were used to validate those which emerged from the administration of the survey questionnaire in the quantitative study. Figure 4 is a schematic representation of the procedures which the researcher followed to triangulate the main findings from both phases. From the results, it can be concluded that the mobile money payment and receipt services have impacted on the financial performance of the SMEs in Douala, Cameroon after they had begun to use the platform.

**Figure 4.** Triangulation of findings (SMEs: small and medium-sized enterprises. Source: authors).

#### **5. Conclusions and Recommendations**

Over the past decade, Mobile Money has improved financial inclusion in several developing countries but has also improved the way of life of many households and the business operations of many SMEs. The Mobile Money services merits could be used to address some of the difficulties faced by the SMEs in Douala, Cameroon. Mindful of these advantages and the growing uptake of the platform in Douala, the researchers investigated the impact of the mobile money payment and receipt services on financial performance. Taken collectively, the independent variables predicted 73% of the variance in sale turnover of the SMEs that participated in the study after they had begun to use Mobile Money services. In line with the literature, it can be concluded that the adoption of Mobile Money services exerted a significantly positive influence on the financial performance of the SMEs in this study. A finding which could plausibly be generalised at least to the two markets in Douala in Cameroon in which the study was conducted.

Considering the principal sources of difficulty which the participants identified with respect to the effective running of their businesses, and the perceived role of SMEs in economic growth, one would recommend firstly that the SMEs in Douala should make full use of Mobile Money services given its potential to improve financial performance. Secondly, that the government should investigate the feasibility of promulgating laws that would make Mobile Money service providers licensed financial institutions, as this would significantly reduce their operating costs and enable them to make their services more accessible to users. The partnerships which Mobile Money service providers have with commercial banks at present entail considerable expense and prevent them from making their services more affordable. Thirdly, the paper recommends that the government should encourage SMEs to make Mobile Money transactions through appropriate tax incentives. Finally, the government, through the Ministry of Small and Medium-sized Enterprises, Social Economy, and Handicrafts as one of its chief regulators, should provide support to the SME sector in the form of policies that facilitate an environment which is conducive to economic growth. Commercial banks must be encouraged to provide financial assistance to SMEs, even if doing so necessitates the government assuming responsibility for loans which are made to SMEs which have been assessed as being viable.

#### **6. Limitations of the Study and Future Research Directions**

One of the principal limitations concerned the researcher being obliged to rely upon the subjective assessments of the respondents to the survey questionnaire and interviewees on the influence of Mobile Money on their turnover. The accuracy of the findings would improve significantly if the researcher had access to informal financial records and could observe the levels of turnover from the numbers of sales which were transacted before and after the adoption of Mobile Money in a longitudinal study. Also, it needs to be conceded that reducing the initial sample size of 384, which had been calculated through the formula of Cochran [59], to 285 could have altered the findings. Using growth for turnover as the sole determinant of the increased financial performance of the SMEs without investigating the influence of other factors could have constituted another limitation. The statistical analysis of the data did not include a normality test, and only significant variables were used in the inferential analysis. The degree to which the findings could be generalised to other target populations could also be limited unless the business practices of the SMEs concerned are very similar to those in the two markets in Douala.

To enhance the generalisability and credibility of the findings, further investigations on the topic in Cameroon should prioritise the use of quantitative methodologies. Additionally, for future research, turnover should be based upon recorded sales, even if the methods which are used to record sales are informal. Given that the paper relied on the use of cross-sectional data, further studies should benefit from the longitudinal survey. Finally, as the independent variables which were identified as contributing to variance in sales turnover were found to predict of the order of 73% of the variance, those which could predict the remaining 27% could be investigated in future studies.

**Author Contributions:** F.S.G.T. and R.K.T. designed and performed the study. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A**


#### **Table A1.** Reliability scores for each variable.

Source: authors.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Do CSR Ratings A**ff**ect Loan Spreads? Evidence from European Syndicated Loan Market**

**Danilo Drago 1,\* and Concetta Carnevale <sup>2</sup>**


Received: 14 August 2020; Accepted: 9 September 2020; Published: 16 September 2020

**Abstract:** We investigate whether corporate social responsibility (CSR) ratings affect the syndicated loan spreads paid by European listed firms. By performing ordinary least squares (OLS) pooled regressions on a sample of 1101 syndicated loans granted to European companies, we find evidence that borrowers' CSR ratings have a significant impact on loan spreads. However, the relationship between CSR ratings and loan spreads is quite complex. Low CSR-rated firms pay higher loan spreads than better CSR-rated firms, but high CSR ratings are not always rewarded by lenders. The benefits of a high CSR rating level are significant only for firms located in countries that pay great attention to sustainability issues. Overall, our work provides a key to reconciling the mixed results obtained in the empirical literature, as we find evidence of a significant lack of homogeneity within the European Union countries regarding the relationship between CSR performance and the cost of debt financing.

**Keywords:** corporate social responsibility; CSR rating; bank loan spread; European syndicated loan market

#### **1. Introduction**

The topics of corporate social responsibility (CSR) and sustainable growth have been investigated by the academic research for many years. Recently, the aforementioned themes have also entered the agenda of the policymakers, at least in the European Union. In late 2016, the European Commission established the EU High-Level Group on Sustainable Finance (HLEG) to help develop an overarching and comprehensive EU roadmap on sustainable finance. In its final report, released in January 2018, the HLEG argued that "the primacy of banks among lenders in assessing the credit risk of individual loans makes them particularly important for financing the origination of sustainable assets and for lending in support of the transition to a more sustainable economy" [1] (p. 67). In addition, in May 2020, the European Banking Authority (EBA) released a document defining the guidelines on loan origination and monitoring. In this document the EBA states that institutions "should take into account the risks associated with ESG factors on the financial conditions of borrowers" [2] (p. 26). A final unequivocal statement of political will is provided by the priorities set out by Ursula von der Leyen in the *Political guidelines for the next European Commission 2019–2024* [3].

Given these premises, we aim to investigate the relationship between the cost of bank loans and corporate social performance (CSP) in the European context. Understanding the attitude of banks toward the CSP of their borrowers allows us to draw important implications both for policymakers and for firms' managers. If banks do not consider CSP in the assessment of borrowers' credit risk, or, even worse, if banks apply greater spreads to borrowers with better CSP, then companies will be less inclined to bear the costs of CSR engagement. On the other hand, if banks reward borrowers who exhibit better CSP with lower loan spreads, then the managers of borrower firms will receive clear

indications about the investment policies to be set in the long term, and at the same time, they will be able to more easily justify the increase in costs originating from CSR engagement.

Our study aims to verify whether CSR ratings, a measure of CSP provided by a specialized rating agency, affect syndicated loan spreads charged to European listed firms. Our results suggest that banks consider CSR rating levels when they assess borrowers' creditworthiness. We find that CSR ratings are on average negatively related to loan spreads. However, by decomposing the average effect, we show that this relationship is more complex. Unlike existing studies investigating the European context [4–6], we explicitly examine potential nonlinearities in the relationship between CSR rating and the cost of debt financing, and we find that the country's environmental, social and governance (ESG) performance significantly affects the CSR rating–loan spread link. In high ESG-rated countries, a firm's loan spread declines as its CSR rating improves. In low ESG-rated countries, there seems to exist a U-shaped relationship between CSP and cost of debt financing: both high and low CSR-rated firms pay higher spreads than those with median CSR ratings. This implies that high CSR ratings do not automatically lower firms' credit risk. Our results also hold considering potential endogeneity issues, lender characteristics, borrower's credit quality, and crisis periods.

Our work fits into the literature that investigates the relationship between CSP and corporate financial performance (CFP). The literature on this topic is large (for a review, see Brooks & Oikonomou [7]), but only a limited stream of studies investigates the impact of CSP on firms' credit risk. Moreover, most available studies are focused on US firms [8–11] and find weak evidence that greater CSP leads to decreases in credit risk and credit spreads. The results obtained by examining US firms cannot be mechanically extended to European capital markets. Looking at the problem through the lens of institutional theory and the Varieties of Capitalism (VoC) theory, we know that the institutional context affects CSR activities and CSP [12]. Furthermore, in the US market-based system, companies are able to raise capital through large and liquid securities markets. In the credit-based system of Continental Europe, companies face rather thin capital markets and meet their financial needs mainly through bank loans. A study by the European Commission revealed that only 1498 of the more than 50,000 companies with assets over 10 million euros had access to the bond market [13]. Given that the European financial system is predominantly bank-based, banks are the key player to consider when investigating whether a borrower's CSR commitment can offer any contribution to the reduction of credit risk. The few articles investigating the relationship between CSP and credit risk in the European market are focused on the bond market [4,5]. Therefore, the relationship between CSP and the cost of bank debt in the European context is still largely unexplored.

To the best of our knowledge, the only study that includes European companies in an international sample of syndicated loans is Hoepner et al. [6], although their sample includes only 195 loans granted to European borrowers. The authors claim to find no conclusive evidence that firm-level sustainability influences the interest rates charged to borrowing firms by banks. Moreover, their findings do not support the view that the country's sustainability rating moderates the CSP–loan spread link. Unlike Hoepner et al. [6], we find evidence of a significant relationship between CSP and the cost of bank debt, although this relationship is not uniform throughout the European countries, but is conditional on the borrower's country ESG rating. In line with Stellner et al. [5], we claim that the benefits resulting from CSR investments are context driven. A detailed knowledge of the cross-country differences affecting the CSP–CFP link is particularly important to understand how banks reward CSP when they evaluate the creditworthiness of their borrowers.

In addition, banks not only play a dominant role in the European financial system, but they also differ significantly from other economic agents. Given their continuous monitoring activity and long-term customer relationship with firms, banks are considered "quasi-insiders" of firms. Therefore, banks can assess firms' creditworthiness better than other entities. This element allows us to analyze how more informed lenders evaluate borrowers' CSP.

We emphasize that the loan market may incorporate CSR information differently than the bond market. There has been a remarkable increase in sustainable investing. Sustainable investing is an investment approach that considers CSR-related factors in portfolio management. According to the Global Sustainable Investment Alliance [14], investors representing over half of all professionally managed assets in Europe adopt some form of screening based on sustainability filters. Given the increase in sustainable investing, bonds issued by socially responsible companies may be, at least in part, purchased by investors who do not make their investment decisions on the basis of mere economic convenience. Since many fund managers must consider the CSR commitment of firms issuing bonds in order to comply with their investment mandate, firms engaged in CSR could have a stable advantage in terms of greater demand in the bond market. In contrast, this benefit could be more uncertain in the syndicated loan market because lender banks are not contractually required to consider firms' CSR engagement. Thus, in line with the existing literature [8], we assume that banks have no social agenda to promote.

Our study offers several contributions. First, we contribute to the debate on the CSP–CFP relationship by adding insights about the information that lenders take into account when they decide which loan spread to charge to their borrowers. Our findings provide support for a significant relationship between CSP and CFP. However, CSP remains a second-order determinant of loan spreads compared to credit ratings and other financial and accounting variables.

Second, we provide empirical evidence of the CSP–credit risk link in the European market, which is characterized by a different institutional context with respect to the markets investigated until now by the existing literature, mainly focused on the US market. Our study fills this gap and provides evidence that CSR ratings significantly affect the cost of bank debt.

Third, our results highlight the importance of cross-country heterogeneity to depict a comprehensive picture of the CSP–CFP relationship in the European context. We document that the European Union cannot be considered as a homogeneous area, because we find that the cross-country differences in the attention to ESG issues affect the relationship between CSP and the cost of debt financing. These results are undocumented for European firms in the CSR literature, and they are partially at odds with findings from the U.S. context [15].

Finally, we provide evidence of significant nonlinearities in the CSP–CFP relationship. These findings are consistent with the view that country awareness and sensitivity toward ESG issues may be able to constrain companies from making excessive and wasteful investments in CSR.

Overall, our work provides a key to reconcile the contradictory results obtained from literature with reference to the European firms.

The remainder of the paper is organized as follows. Section 2 reviews the related literature. Section 3 presents the research hypotheses. Section 4 describes the data and methodology. Section 5 shows our main results. Section 6 refers to the robustness checks and additional results. Section 7 provides a discussion of our results. Finally, Section 8 concludes.

#### **2. Literature Review**

There has always been some skepticism among practitioners and researchers about the value of CSR. For practitioners, a signal of a changed attitude about the role of CSR comes from the following [16]: "Beyond the attempt to deceive customers and regulators, the [Volkswagen] scandal also highlights the failure of traditional valuation models—such as discounted cash flow—to capture the full range of risks companies face today. It also underlines the potential benefit of assessing companies with alternative data sets that highlight environmental, social, and governance (ESG) signals, flagging risks that traditional analytical tools aren't designed to identify".

Although the skepticism toward CSR has not completely disappeared, there is now a growing body of literature that identifies numerous positive effects of CSR commitment. CSR investments may become competitive advantages for firms, because they allow companies to build internal resources by improving their reputation and customer loyalty [17,18]. By engaging in CSR, firms can improve relationships with their stakeholders [19], resolve conflicts between various groups of stakeholders [20], and be less exposed to legal, reputational, and regulatory risks relating to controversial or irresponsible activities [9].

At the same time, CSR investments may be perceived as a signal of superior management skills [21]. CSR may likely lead to better economic and financial performance, because it is also connected to trustworthiness, integrity, non-opportunistic behavior, and the moral character of a firm [6].

The existing literature has proposed two different hypotheses explaining the CSP–CFP link: the *risk mitigation view* and the *overinvestment view* [8]. Under the *risk mitigation view*, superior CSP is regarded as a factor that improves the risk profile of a company. Companies that invest in CSR are able to strengthen their relationships with key stakeholders and to build internal resources and intangibles that provide stability and a buffer in times of downturn and should result in lower cash flow volatility. The better risk profile and the greater ability to repay the principal at maturity are rewarded by lenders with a lower spread charged to the borrower company [5]. Under the *overinvestment view*, investors regard investments in CSR as a waste of scarce resources. Excessive costs for handling the various relations with a high number of stakeholders may increase complexity and reduce profitability, leading to higher borrowing costs.

A growing literature focuses on the effect of only one dimension (environmental, social, or governance) of CSP on credit risk. See, for example, Nandy and Lodh [22]; Chava [23]; Kim et al. [24]; Cui et al. [25]. These studies are linked to our work, but we adopt a broader perspective, investigating the relationship between CSR ratings (the overall CSR performance) and loan spreads. Several studies find empirical evidence supporting the *risk mitigation hypothesis* by examining the impact of CSP on firms' financing costs. El Ghoul et al. [26] document that the cost of equity is lower for US firms with better CSR scores. In examining the impact of CSP on bond spreads and the ratings of US firms, Oikonomou et al. [9] show that CSP is negatively but weakly related to systematic firm risk, and that corporate social irresponsibility is positively and strongly related to financial risk. Jang et al. [27] find that higher ESG scores can help lower the cost of funding for the bond issuers of relatively small Korean firms. Salvi et al. [28] investigate the international bond market. They find that superior CSP strengths are associated with lower credit spreads, while a higher number of CSP-related controversies leads to an increase in the cost of corporate bonds. Truong and Kim [29] analyze the U.S. credit default swap market and show that CSR activities reduce credit risk in the long run more than in the short run. Gangi et al. [30] find that CSP has a significantly negative influence on the firms' risk of financial distress. In line with the risk mitigation view, Bae et al. [11] provide evidence that CSR matters to the pricing of US loan contracts, and that the absence of scrutiny by credit rating agencies exacerbates the lenders' negative view in case of poor CSP. Bouslah et al. [31] find that the impact of CSR dimensions on firms' risk is not uniform, and that, in general, the relation between firms' risk and CSR strengths and concerns is more significant for more transparent firms (included in the S&P 500 index) than for more opaque companies (not included in the S&P 500 index). Ge and Liu [32] show that the disclosure of better CSP leads to lower yield spreads. In addition, they document that firms with weaker CSP do not pay significantly different yield spreads than firms that do not disclose CSR information. Stellner et al. [5] find only weak evidence that superior CSP results in reduced credit spreads in the European corporate bond market. Moreover, they show that the relationship between CSP and credit risk is conditional on a country's ESG performance.

Compared to the above mentioned studies, the *overinvestment hypothesis* offers an alternative view, drawn from the agency theory. Under this view, a higher CSR engagement pushes firms' investments over the optimal level. From the shareholder perspective, by engaging in CSR activities, firms divert resources from the maximization of shareholder wealth [33,34]. At the same time, CSR activities may increase firms' fixed costs and the volatility of earnings, leading to an increase in firms' default risk [35].

In addition, given the existence of principal–agent conflicts of interest, managers can use CSR activities to improve their own reputation at the expense of shareholders [36]. In this view, CSR investments can be assimilated to other agency costs, such as the purchase of unnecessary corporate jets [8,37]. Other researchers, drawing from neoclassical economic theory, argue for a negative relationship between CSP and CFP. These authors contend that responsible firms are at a competitive disadvantage compared with their unresponsive peers [38–40].

Consistent with the *overinvestment hypothesis*, Menz [4] finds that CSP is positively related to European corporate bond spreads, but this relationship appears only weakly significant. Goss and Roberts [8] show that CSP leads to an economically modest decrease in loan spreads applied to US public firms. However, they find evidence of a positive relationship between CSR investments and loan spreads applied to low-quality borrowers, because the agency costs associated with sustainable investments are greater for these firms. Baran and Zhang [41] show that the yields of newly issued bonds are greater for firms included in the KLD 400 Index. Hoepner et al. [6] do not detect a significant link between CSP and syndicated loan spreads and document that particular dimensions of CSR even appear to lead to greater loan spreads. Finally, Bae et al. [15] find evidence of a non-linear effect of CSR investments on debt financing costs in a sample of US firms.

Our study presents important elements of novelty compared to previous studies, which are focused on the European context [4,5] or investigate some international samples that include European companies [6,23]. Unlike the above mentioned studies, we hypothesize that the European context is not a homogeneous area, and we prove the existence of nonlinearities in the relationship between CSP and loan spreads in the European area.

#### **3. Hypothesis Development**

#### *3.1. Do CSR Ratings A*ff*ect Loan Spreads?*

The risk mitigation view and the overinvestment view offer the theoretical background to verify whether CSR ratings affect the cost of bank loans.

Lenders take into account potential risks that may negatively affect the borrowers' financial performance. In this respect, lenders may be concerned about the likelihood that CSR-related issues (e.g., a corporate scandal or a negative environmental event) increase default risk and jeopardize the ability of the borrower to repay his debts. Under a broader perspective, as Bae et al. [11] point out, CSR engagement may reduce conflicts of interests between managers and stakeholders. If the conflict-resolution hypothesis holds, then CSR engagement reduces agency costs and conflict of interests among various stakeholders, including lenders. Thus, under the risk mitigation view, we expect that:

#### **Hypothesis 1a (H1a).** *CSR ratings are negatively related to loan spreads*.

On the contrary, lenders may consider a firm's CSR engagement from the perspective of a principal–agent relationship between managers and shareholders: CSR investments waste corporate resources and thus make borrowers more vulnerable to adverse economic conditions. The competitive disadvantage hypothesis (neo-classical economic theory) reaches the same conclusions. Under the previous views, lenders charge higher spreads to high CSR performers borrowers. If this is the case, consistent with the overinvestment view, we can propose an alternative hypothesis:

#### **Hypothesis 1b (H1b).** *CSR ratings are positively related to loan spreads*.

However, the previous hypotheses could only provide a first approximation of the actual relationship between CSR ratings and loan spreads. Bae et al. [15] combine the risk mitigation hypothesis with the overinvestment hypothesis, suggesting a non-linear relationship between CSP and loan spreads. Building on previous theoretical studies, they hypothesize an optimal level of CSR investments that maximize profits, while also satisfying the demand for CSR of the other stakeholders. The optimal level of a firm's CSR investments is that required to fully insure the firm's risky assets against loss, so CSR investments beyond this level would impose additional costs without producing any insurance benefits. The authors provide evidence of a U-shaped relationship between CSP and debt financing costs for a sample of U.S. bank loans. Similar findings are reported by Ye and Zhang [42] for Chinese firms.

We too hypothesize a nonlinear relationship between CSP and loan spreads, but we assume a different relationship from that of Bae et al. [15]. Several studies find evidence that CSP has a mitigating effect on stock price crash risk [43] and on downside risk [44,45]. From a theoretical point of view, the previous literature argues that CSR investments can reduce a firm's risk exposure through insurance-like protection by generating moral capital among stakeholders. The creation of moral capital (and other intangible, internal resources) acts as insurance-like protection when negative events occur, preserving shareholder value. We emphasize that lenders, compared to shareholders, are more averse to downside risk. As a result, lenders may be less willing to penalize high CSP levels than shareholders. In addition, we have already clarified (see Introduction) that the results provided by Bae et al. [15] for the US market cannot be mechanically extended to other institutional contexts. In this respect, further confirmation comes from Utz [43], who examines the predictive power of CSP for both idiosyncratic risk and stock crash risk in an international sample and finds mixed results. In the Asia-Pacific sample, high CSP increases crash risk, in accordance with the overinvestment hypothesis. On the contrary, in the European sample, there is no evidence of a U-shaped relationship between CSP and idiosyncratic risk.

Building on the above-mentioned studies, we may hypothesize that better CSR ratings lower loan spreads but at a decreasing rate. High CSP may increase firms' fixed costs or create a competitive disadvantage. However, the high aversion of lenders to downside risk, together with the existence of specific institutional or cultural factors affecting the European context, may counteract the increase in a firm's fixed costs, preventing a positive relationship between loan spreads and CSR ratings. In light of previous consideration, we propose an additional hypothesis:

**Hypothesis 1c (H1c).** *Loan spreads are not a strictly decreasing function of CSR ratings: as the CSR rating increases, the loan spread function should first be decreasing and then should become approximately flat*.

In any case, to further validate our hypothesis, we will also control for a potential U-shaped relationship in our sample.

#### *3.2. Does Country ESG Performance A*ff*ect the CSR Rating–Loan Spread Link?*

The institutional theory argues that the national institutional and economic environment influences the likelihood that companies will assume CSR compliant behavior [46,47], and that variation in CSP across firms is explained by variation in national-level institutions [12]. Cai et al. [48] document the role of other country factors, besides national institutions, that explain CSP, such as differences in stages of economic development, the cultural dimension, factors associated with the political system (e.g., corruption, civil liberty, and political rights), as well as the education and labor system characteristics. Following this line of reasoning, Hoepner et al. [6] outline that issues such as climate change, resource scarcity, population growth, and ageing have deep economic repercussions, and that ESG macro-themes have a growing importance in the valuation of every asset class and type of financial contract. They find that a higher country sustainability rating is associated with lower costs of bank loans and argue that the sustainability framework of the home country "act[s] as a shield for the borrower firm, protecting it from the operational and reputational hazards occurring from systemic social and environmental challenges and, ultimately, reducing its default risk" [6] (p. 161). Stellner et al. [5] show that the relationship between CSR engagement and EU firms' credit ratings and bond spreads depends significantly on the CSP of the country where the company is established. They argue that CSP leads to lower credit risk only if the CSR efforts of firms are rewarded in the environment in which they are embedded. In particular, the authors find that greater CSR efforts lead to greater benefits for companies whose CSP mirrors that of their home country.

Given the potential effect of countries' sensitivity to CSR issues, we hypothesize that the impact of CSR ratings on syndicated loan pricing may be affected by the home country ESG rating of the borrower. Lenders should reward borrowers whose high CSR rating mirrors the home country ESG performance. Conversely, in countries with low ESG performance, high CSR ratings may bring lower benefits to the borrower or may be associated with higher borrowing costs. Moreover, low CSR-rated firms should pay higher loan spreads regardless of the country ESG rating. Indeed, potential lenders could hardly ignore issues related to poor CSP (e.g., problematic relationships with consumers, employees, and other stakeholders) both in high and low ESG-rated countries.

Overall, we expect that in high ESG-rated countries, the loan spread is a decreasing function of CSR rating. On the contrary, in low ESG-rated countries, we expect to find evidence consistent with a U-shaped relationship between loan spread and CSR rating. Therefore, we formulate our second hypothesis:

**Hypothesis 2 (H2).** *In high ESG-rated countries, firms pay lower loan spreads as their CSR rating improves. In low ESG-rated countries, high or low CSR-rated borrowers pay higher loan spreads compared to median CSR-rated borrowers*.

#### **4. Data and Methodology**

#### *4.1. Sample and Data*

Our sample consists of syndicated loans granted to listed non-financial firms established in EU member states during the 2006–2015 period.

We use LPC's DealScan database to collect information on individual loans, including: the loan closing date, the loan spread over Libor (incorporating any annual or facility fees paid by the firm), maturity, seniority status, purpose, and type. We also retrieve from LPC's DealScan database the information on the borrower, including its sector of activity, and the lending syndicate, including the identity and the role of banks in the loan syndicate.

For each firm, we retrieve CSR ratings from Thomson Reuters ASSET4. The ASSET4 database covers more than 6000 companies around the world, enabling us to investigate the European context. ASSET4 ratings have a reputation for being among the most diligent and trustworthy sources of CSR data [5,30,43]. ASSET4 assigns a score to each company considering four pillars: environmental, social, corporate governance, and economic. These four pillars have approximately 750 individual data points, which are combined into 280 key performance indicators (KPIs). Then, these KPIs are structured into 18 categories within the four pillars. ASSET4 provides a score for each pillar and an equal-weighted rating, which indicates the overall CSR score. Each score is calculated by equally weighting and z-scoring all underlying data points and comparing them against all companies in the ASSET4 universe. The final score is expressed as a percentage and is therefore a relative measure of performance.

We retrieve data on ESG country ratings from Bloomberg. Bloomberg provides an overall score for more than 170 countries and an individual score in the dimensions of environmental, social, strategic governance and economics that matches the four categories provided by ASSET4.

We use Datastream to collect information on firms' balance sheets. We match firms in LPC's DealScan to Datastream, using the company name and ISIN code, to extract firms' accounting information. After the matching, our dataset consists of 1727 loans granted to 483 firms. It is worth noting that, in our sample, approximately 60% of borrowers have a CSR rating, while only 40% of borrowers have a credit rating. This comparison highlights the relevance gained by CSR ratings for European listed firms.

Finally, after excluding from our dataset the companies without a CSR rating, our final sample includes 1101 loans granted to 297 firms.

#### *4.2. Methodology*

#### 4.2.1. Measuring CSP

Measuring CSP is a challenging task involving the assessment of a broad range of economic, environmental, governance, and social factors [49]. Existing studies have adopted a remarkable

variety of different CSP measures [50]. More recently, several studies have measured CSP by adopting the assessments provided by social rating agencies. Following these studies, we measure CSP by employing Thomson Reuters ASSET4 ratings (described in Section 4.1). We are aware that any CSP measure involves unavoidable elements of subjectivity. However, we underline that, in contrast to measures specifically built for a single research work, a rating provides a CSP measure that is public and available to the entire financial community.

According to the economic theory, rating agencies perform at least two main functions: signaling and monitoring. For example, credit rating agencies signal to investors the creditworthiness of the issuer of a financial security (signaling), and, after the security issuance, they continue to monitor the issuer (monitoring). Similarly, social rating agencies signal and monitor CSP.

The literature notes that sustainability commitments are difficult to verify. Consumers, investors, and other external stakeholders are not able to verify the sustainability claims made by companies, because they do not have access to relevant information [51]. Reliable third party institutions, which are able to gather the needed information, may become important players [52].

We do not claim that CSR ratings are the best possible way to measure CSP. More simply, we aim to verify whether CSR ratings affect banks' loan pricing decisions. Again, we can find an analogy between CSR ratings and credit ratings. It is well known that the credit rating is not the only possible measure for assessing creditworthiness, and that the rating alone is not able to explain the level of credit spreads paid by different borrowers. However, it is generally recognized that credit ratings provide the market with economically relevant information. Furthermore, CSR ratings enjoy a significant difference compared to credit ratings. The latter are widely used in financial regulation, and economic agents are in some way obliged to take the credit rating into account in their decisions. In contrast, CSR ratings have not, to date, been subject to regulatory use. Consequently, economic agents can freely decide whether to consider the CSR rating, without being conditioned by regulation.

It must also be recognized that CSR ratings suffer from many shortcomings [53,54]. Windolph [52] highlights several undesirable properties of CSR ratings described in the literature: lack of standardization, lack of credibility of information, bias, lack of transparency, and lack of independence. Chatterji et al. [55] document a surprising lack of agreement across social ratings from six well-established raters. The authors claim that low convergence of social ratings remain even when they adjust for explicit differences in the definition of CSR held by different raters.

These problems do not invalidate our analysis. If our research hypotheses were verified, we could state that the ASSET4 rating provides information relevant for the pricing of the syndicated loans. This result, if proven, would not exclude that different CSR rating measures may provide other relevant information not captured by ASSET4 ratings.

Social raters began releasing their assessments only recently, especially when compared to the longstanding experience of credit rating agencies. Over time, market forces will select rating agencies, allowing only those agencies able to provide economically relevant information to survive [56].

#### 4.2.2. CSR Rating and Loan Spread

To test our research hypotheses, we perform ordinary least squares (OLS) pooled regressions, treating the facilities in each deal as different loans. Consistent with the literature (e.g., [8]), we focus on cross-sectional differences, because most firms have only a few different observations. It is worth noting that we consider a sample of loans observed at their origination. Most firms have received just few loans during our sample period. As we treat the facilities in each deal as different loans, most borrowers often receive multiple loans at the same date, and subsequently they no longer appear in our sample. For the same reason, we have just one observation for several borrowers. Given the sample characteristics, we do not adopt firm fixed effects or other panel techniques. However, we perform additional tests to address self-selection bias and potential endogeneity and reverse causality issues (see Appendix A).

Our base regression model is described in Equation (1):

$$LnSpread\_{i,t} = \beta\_0 + \beta\_1 CSR\_{i,t-1} + \beta\_2 B\_{i,t-j} + \beta\_3 L\_{i,t} + \beta\_4 X\_{i,t-j} + \varepsilon\_{i,t} \tag{1}$$

The dependent variable is the logarithm of the all-in-drawn spread of the loan granted to the *i*-th firm at the loan closing date *t*. Since borrowers are unlikely to receive loan spreads lower than LIBOR, the spread variable may be characterized by a positive skewness. Thus, we use a log-transformed spread to mitigate this potential bias.

The vector *CSR* includes alternative key explanatory variables used in our estimates. To test *H1a* and *H1b*, we use the variable *EW Rating*, which is the CSR equal-weighted rating of the *i*-th firm in the year preceding *t*. A negative and significant coefficient of *EW\_Rating* would imply that banks charge higher loan spreads to firms with lower CSR ratings than firms with higher CSR ratings. The opposite is true if the *EW\_Rating* coefficient proves to be positive and significant.

To test our hypothesis *H1c*, we introduce the following variables: *HighEWRating*, which is a dummy variable that is equal to 1 if the CSR equal-weighted rating of the *i*-th firm in the year preceding *t* is in the highest tertile of the empirical distribution; and *LowEWRating,* which is a dummy variable that is equal to 1 if the CSR equal-weighted rating of the *i*-th firm in the year preceding *t* is in the lowest tertile of the empirical distribution. Then, in Equation (1), we replace *EW\_Rating* with *HighEWRating* and *LowEWRating.* This distinction allows us to verify whether the negative relationship between CSR ratings and loan spreads is not strictly decreasing. To further test *H1c*, we drop *HighEWRating* and *LowEWRating*, and, following Bae et al. [15], we estimate Equation (1) by adding the quadratic term *EW\_Rating*<sup>2</sup> . If the coefficient of this variable proves to be significant, our hypothesis would not be confirmed.

#### 4.2.3. Control Variables

Following the existing literature on syndicated loan spreads, we develop our model by including three vectors of control variables [57,58]. The vector *B* includes *Borrower Variables* to consider firms' accounting information (*Size*, *CashFlow*, *ROS*, *IE\_Revenue*, *Leverage*), market-based data (*MTBV*, *Stock\_StdDev*), and industry (*Industry*). In addition, we consider the role of credit ratings. Previous studies emphasize that CSR-related risk factors may affect credit ratings [5,10,59,60]. If credit ratings fully incorporate CSR-related information that is relevant for lenders, then we will observe a non-significant impact of CSR ratings when we include borrowers' credit ratings in our model. Therefore, we introduce the variable *Risk\_Weight*, which indicates the risk weight assigned to the *i*-th firm under the Basel II standardized approach. To calculate *Risk\_Weight*, we convert the borrower's rating into a risk weight by adopting the weighting scale used in the Basel framework. *Risk\_Weight* takes values from 0 to 1.5. High values of this variable indicate a greater credit risk of the borrower. This approach offers two advantages: (i) it allows us to convert an ordinal risk measure (rating) into a cardinal measure (risk weight), and (ii) it allows us to assign a risk weight (equal to 1) to unrated firms [61].

In Loan Variables, the vector L, we consider loan characteristics: Maturity, Secured, Covenant, Purpose, Type, Seniority, and Loan\_Concentration. Following Goss and Roberts [8], we use Loan Concentration as a proxy for the strength of the relationship between the bank and the borrower. In our robustness tests, we specifically address this point considering the potential impact of relationship banking effects (Appendix A).

Finally, we use the vector *X* to control for the stock index return of the country where the *i*-th firm is established over the three months preceding *t* (*Sov\_Stock\_Ret*), the *i*-th firm's home country sovereign rating (*Sov\_Rating*), the loan reference rate value (*Ref\_rate*), and year dummies (*Year dummies*). In vector *X*, we also include a set of variables to control for country effects. To achieve this goal, we follow the varieties of capitalism (VoC) approach, which suggests that the different models of the economic system adopted by each country affect the national financial market [62–64]. Comparative analysis of

capitalism is based on an identification of a set of key institutional areas: (i) product market competition; (ii) the labour market; (iii) the financial intermediation sector (regarding the financial intermediation sector, the cluster classification considers, inter alia, the following variables: sophistication of financial markets, stock ownership concentration, creditor rights protection, importance of institutional investors, degree of banking concentration, and importance of banks in firms' investment funding) and corporate governance; (iv) social protection; and (v) the education sector. Countries exhibit significantly different features in each of these areas, and institutional complementarities define the different domestic models of capitalism. According to the VoC approach [64], EU countries can be clustered into five different models of capitalism: liberal market economies (*Lme*), liberal-like market economy (*LLme*), coordinated market economies (*Cme*), state-dominated market economies (*Sd*), and hybrid market economies (*Hy*). Therefore, to identify the capitalism model of the *i*-th firm's home country, we introduce a dummy variable for each cluster. In alternative versions of our models, we have replaced VoC dummies with country fixed effects, obtaining qualitatively similar results. Results unreported for space considerations are available from the authors.

In the Supplementary Material (hereafter SM) we report in Table S1 the complete list of variables used in our study and their relative sources.

#### 4.2.4. The Role of the ESG Country Rating

To test our hypothesis *H2* we perform two different regression analyses. First, we drop *EW Rating* in Equation (1), and we reintroduce the dummy variables that divide our sample into tertiles (*HighEWRating* and *LowEWRating*). Then we add the following variables: (i) *HighESGCountry*, which is a dummy variable equal to 1 if the ESG rating assigned by Bloomberg to the *i*-th firm's home country in the year preceding *t* is above the average of the sample, (ii) the interaction variable *HighEWRating*·*HighESGCountry*, and (iii) the interaction variable *LowEWRating*·*HighESGCountry*. In this way, we consider six groups of firms: (1) high-rated firms located in low ESG-rated countries; (2) firms with median CSR ratings located in low ESG-rated countries; (3) low-rated firms located in low ESG-rated countries; (4) high-rated firms located in high ESG-rated countries; (5) firms with median CSR ratings located in high ESG-rated countries; (6) low-rated firms located in high ESG-rated countries. Our benchmark group includes firms with median CSR ratings established in countries with lower ESG rating (group (2)).

Consistent with *H2*, we expect that: (a) firms in group (1) and (3) pay higher spreads than firms in group (2); and (b) firms in group (6) pay higher spreads than firms in group (5), which, in turn, pay higher spread than firms in group (4).

To better investigate potential nonlinearities in the relationship between CSR ratings and loan spreads, we run a second regression analysis splitting our sample into high ESG-rated and low ESG-rated countries. Then, we add in Equation (1) the quadratic term of *EW Rating*. In order to confirm *H2*, we should find that in low ESG-rated countries the coefficient of *EW\_Rating*<sup>2</sup> is positive and significant. In contrast, in high ESG-rated countries, our hypothesis will be confirmed whether (i) the coefficient of *EW\_Rating*<sup>2</sup> is negative and significant and (ii) the interpolating parabola is a decreasing function in the domain of *EW Rating*.

#### **5. Results**

#### *5.1. Sample Characterization*

In Table 1, we report the distribution of loans included in our sample by country. We also indicate the percentage distribution, the mean all-in spread, and the mean amount. We observe that firms included in our sample are established in 17 EU member states.


**Table 1.** Distribution of syndicated loans by country.

Table 2 shows the summary statistics for the main variables used in the regression models.


**Table 2.** Summary statistics.

Notes: The sample consists of 1101 loans granted to listed non-financial firms established in EU member states during the 2006–2015 period.

We checked the correlations among variables, and we can affirm that the correlations do not represent a concern for our estimates (please see Table S2 in Supplementary Materials).

#### *5.2. The Impact of CSR Ratings on Loan Spreads*

In Table 3, we report the estimates obtained by testing our alternative hypotheses *H1a* and *H1b* (the complete results are reported in Supplementary Materials in Table S3). We find that the CSR rating level of the borrower has a significant impact on loan spreads. In particular, the negative sign of the *EW Rating* coefficient suggests that an increase in the CSR rating of 10 scores reduces by about 4.2% the average loan spread applied to borrower firms (our study is focused on the overall CSP, measured by the CSR rating. However, we also verified whether firms' scores in each CSR pillar (economic, environmental, social, and governance) have a different impact on loan spreads. The results obtained for each pillar do not significantly differ from those obtained for *EW\_Rating*. Results are reported in Appendix B). Therefore, consistent with *H1a*, our results suggest that lenders: (i) take into account CSP when they assess borrowers' creditworthiness and (ii) seem to positively evaluate borrowers' CSR efforts.


**Table 3.** The impact of CSR ratings on loan spreads.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

We find plausible results for our control variables. Regarding the variables included in the vector *B* (*Borrower Variables*), size, cash flow, burden of interest expenses, leverage, market-to-book value, and the borrower's stock return have a significant impact on loan spreads. The signs of the coefficients are those expected and in line with the literature. We also observe that greater values of *Risk\_Weight* lead to greater spreads. Since *EW\_Rating* remains also significant controlling for credit rating levels, we can confirm that credit ratings do not fully include CSR-related information and that lender banks consider CSR information also in the presence of credit ratings.

Regarding the variables included in the vector *L* (*Loan Variables*), we note that the priority structure, as expected, has a highly significant impact on the loan spread. In addition, collateral and covenant clauses are associated with greater spreads, because these clauses are generally included in loan contracts for riskier borrowers.

For the third vector of variables (*X*), we note that higher sovereign ratings lead to lower loan spreads. Our results show that, compared to our control group (*CME*), borrowers belonging to the LME countries paid greater loan spreads. This finding is consistent with the VoC literature, which agrees that in the financial markets of LME countries, competitive pressures are higher and that financial transactions are priced according to purely market mechanisms. In contrast, in CME countries, competitive pressures are moderate, financing channels are based mostly on informal relations and on reputational factors, and the relations between banks and companies tend to be long-lasting. The coefficients of the remaining clusters (*LLme*, *Hy*, and *Sd*) are not significant.

Overall, notwithstanding the statistical significance of *EW\_Rating*, we emphasize that the importance of CSR ratings in the syndicated loan pricing process appears relatively limited after controlling for firm and loan characteristics. When we remove the variable *EW\_Rating* from our model, the adjusted R-squared decreases from 67.6% to 66.7%, indicating that the marginal explanatory power

of the borrower's CSR rating level is approximately equal to 0.9%. By comparison, if we remove *Risk\_Weight* from our model, the adjusted R-squared declines from 67.6% to 63.9%, which indicates a marginal explanatory power of the borrower's credit rating level of 3.7% (results unreported for space considerations are available from the authors). Therefore, the incremental explanatory power of CSR ratings is approximately a quarter of that of credit ratings. This result suggests that CSP is considered by lenders, but it remains a second-order determinant of loan spreads compared to credit ratings and other financial and accounting variables.

Overall, our results are consistent with *H1a.* However, as we anticipated in Section 3.1, the actual relationship between *Ew\_Rating* and loan spread could be more complex, so our first regression model may not tell us the whole story.

#### *5.3. High and Low CSR Ratings*

In this section, we report the results obtained for our hypothesis *H1c.* Table 4 shows in column 1 the results of Equation (1) obtained by replacing *EW\_Rating* with *HighEWRating* and *LowEWRating*. In this case, the control group consists of firms with median CSR ratings.


**Table 4.** The different effects of high and low CSR ratings on loan spreads.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

We observe that *LowEWRating* has a significant positive impact on loan spreads, while the coefficient of *HighEWRating* is not significant. These findings are consistent with *H1c*, confirming the diminishing marginal benefits of CSR ratings, and offer some additional insights to *H1a*. Banks charge to low-rated firms a loan spread that is 14% higher on average than that applied to those with better scores. In contrast, high-rated firms do not benefit from a reduction in loan spreads significantly greater than firms with median CSR ratings.

Column 2 shows the results obtained by adding, in Equation (1), the quadratic term *EW\_Rating*<sup>2</sup> . As the coefficient of this variable is not significant, we can exclude that the relationship between CSR ratings and loan spreads is quadratic.

#### *5.4. ESG Country Sensitivity*

In this section, we investigate whether the country ESG rating moderates the impact of firm's CSR rating on loan spread. To this end, we performed a first regression analysis by identifying six groups of firms (Section 4.2.4): groups (1)–(3) include firms located in low ESG-rated countries with, respectively, high, median, and low CSR rating; groups (4)–(6) include firms located in high ESG-rated countries with, respectively, high, median, and low CSR rating.

#### *Sustainability* **2020**, *12*, 7639

Table 5 shows the results of Equation (1) estimated by introducing the variables described in Section 4.2.4. In the reported results, firms with median CSR ratings located in low ESG-rated countries (group (2)) are the control (omitted) group.


**Table 5.** The impact of CSR ratings on loan spreads considering ESG country ratings.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

Our findings depict a complex relationship between the CSR rating and the loan spread when the ESG rating of a firm's home country is taken into account.

First, we examine the spread charged to firms located in low ESG-rated countries (group (1)–(3)). The positive coefficient of *HighEWRating* indicates that better CSR ratings are associated with greater loan spreads if the borrower is located in a low ESG-rated country. Firms in group (1) pay 11.7% more than those in group (2). In addition, the positive coefficient of *LowEWRating* indicates that the spread applied to firms with lower CSR ratings in low ESG-rated countries (group (3)) is 14.3% greater than that charged to companies in the control group (group (2)).

Our findings suggest that in low ESG-rated countries, both high and low CSR-rated firms pay higher spreads than those with median CSR ratings. These results are consistent with the hypothesis suggesting a U-shaped relationship between CSP and the cost of debt in countries less interested in ESG matters.

Second, we analyze the cost of syndicated loans for firms located in high ESG-rated countries (groups (4)–(6)). Given that the coefficient of *HighEWRating*·*HighESGCountry* is significant and negative, the average loan spread applied to high-rated firms located in high ESG-rated countries (group (4)) is 15% (0.117–0.267 using the estimates of Table 5) lower than that charged to firms in group (2) (our control group). In contrast, we observe that the coefficient of *HighESGCountry* is not significant. This implies that the loan spread charged to firms with median CSR ratings established in high ESG-rated countries (group (5)) is not statistically different than that applied to firms with median CSR ratings located in low ESG-rated countries (group (2)). Finally, since the coefficient of *LowEWRating*·*HighESGCountry* is not significant, the spread applied to low-rated firms in high ESG-rated countries (group (6)) is not statistically different than that applied to low-rated firms in low ESG-rated countries (group (3)). Therefore, the cost of syndicated loans for firms in group (6) is about 14% greater than that for firms in the control group (group (2)).

Summing up, in high ESG-rated countries: (a) the loan spread charged to firms in group (6) is higher than that charged to firms in group (2), which pay a loan spread statistically similar to that charged to firms in group (5); (b) the loan spread charged to firms in group (4) is lower than that charged to firms in group (2) and to firms in group (5) also. These results suggest that in high ESG-rated countries, there is no evidence of firms' overinvestment problem, since in these countries the loan spread declines as the CSR rating improves.

To control for potential nonlinearities in the relationship between CSR rating and loan spread, we run additional estimates by splitting our sample into two sub-samples: high ESG-rated countries and low ESG-rated countries. Then, we estimate Equation (1) by introducing the quadratic term *EW\_Rating*<sup>2</sup> . Results reported in Table 6 allow us to provide additional insights. We find evidence of significant nonlinearities in both sub-samples. However, our results depict a divergent relationship between the CSR rating and the loan spread in the two sub-samples. Indeed, we observe that the sign of the quadratic term coefficient is negative for the first group (column 1) and positive for the second one (column 2), suggesting that the relationship between the CSR rating and the loan spread can be described by a function that is concave downward for high ESG-rated countries and concave upward for low ESG-rated ones. These results explain why we were not able to detect a U-shaped relationship between CSP and CFP when we investigated our whole sample (see Section 5.3 above).


**Table 6.** Evidence of nonlinearities in the impact of CSR ratings on loan spreads in high and low ESG-rated countries.

*Notes:* Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

Figure 1 displays the relationship between the predicted values of *LnSpread* (based on the coefficients reported in Table 6) and *EW\_Rating* for both groups of countries. In high ESG-rated countries we observe that firms with a low CSR rating, between 0 and about 33 (the maximum value of the function based on the coefficient reported in column 1 of Table 6), pay approximately the same spread. For firms with a score higher than 33, the loan spread declines as the CSR rating improves, consistent with the risk mitigation view.

In contrast, in low ESG-rated countries, there is a U-shaped relationship between CSR ratings and loan spreads. The spread on loans declines as the firm CSR rating improves until an optimal level of the CSR score, equal about to 74 (the minimum value of the function based on the coefficient reported in column 2 of Table 6). After this threshold, the loan spread rises as the CSR rating improves, implying that there is evidence of firms' overinvestment problem in this group of countries.

Evidence obtained from the cross-country analysis allows us to highlight the role played by the national institutional context in shaping the link between CSP and CFP. Consistent with the institutional perspective, companies mirror their domestic institutional environment by reflecting the actions of the government, market, and civil society. Financial rewards for CSP are linked to the types of corporate behavior sought by society.

In low ESG-rated countries, our findings are consistent with the existence of an optimal level of CSP from the risk mitigation perspective. Lenders positively evaluate firms' engagement in CSR up to the optimal level, because it mitigates firms' exposure to substantial legal, reputational, operational, and financial risks. Hence, CSR engagement would serve as an insurance mechanism against harmful, risk-inducing events. Beyond that optimal level, lenders evaluate the borrower's commitment in CSR

as a waste of the company's resources, because the proactive attitude of companies toward CSR issues does not respond to the requests made by national institutions or by the community.

**Figure 1.** The relationship between the predicted values of loan spreads and CSR ratings in high and low ESG-rated countries.

The previous findings are consistent with those reported by Bae et al. [15], who claim to find a U-shaped relationship between CSR investments and the cost of bank loans in the US context. However, the Bae et al. [15] results do not hold in European high ESG-rated countries. In these countries, lenders positively evaluate CSP as a factor that, all other things being equal, reduces the borrower's riskiness. This positive assessment persists even when the borrower shows high CSP levels. High CSP values are not perceived by banks as a waste of resources. This means that there is no evidence of overinvestment in high ESG-rated countries, where the proactive attitude of companies toward CSR is not penalized.

Our findings can be explained by considering that in countries that show high sensitivity toward CSR issues, companies are encouraged by the institutional context and by the community to pursue a CSR engagement that outpaces the mere risk coverage perspective. At the same time, the high level of awareness and sensitivity of the community toward sustainability issues reduces the risk that management may invest in CSR only for its own interest to improve its own reputation at expense of shareholders.

Overall, our results suggest that (i) the European Union cannot be considered as a homogeneous area, because the cross-country differences in the attitude toward ESG issues affect the relationship between CSR ratings and the cost of debt financing; and (ii) consistent with *H2*, the benefits of high CSR ratings are associated with lower loan spreads only if the borrowers' CSR efforts are rewarded in the environment in which they are embedded. Moreover, in high ESG-rated European countries, the relationship between CSP and the cost of debt financing is consistent with the risk mitigation view; (iii) in low ESG-rated European countries, there seems to be an optimal level of CSR investments. Hence, firms with very high or low CSP are subject to a higher cost of debt, compared to firms with median CSP.

Finally, estimates reported in Tables 5 and 6 allow us a better understanding of the results obtained for the whole sample (hypothesis *H1c*). The relationship observed for the entire sample (first decreasing and then approximately flat) is the result of the mixed impact that the CSR rating exerts on the cost of debt in the different European Union countries: (a) in high ESG-rated countries firm's loan spread declines as the firm CSR rating improves; (b) in low ESG-rated countries there is a U-shaped relationship between CSR rating and loan spread. To the best of our knowledge, the previous results are undocumented for European firms in the CSR literature.

#### **6. Robustness Checks and Additional Results**

In Appendix A, we address potential endogeneity issues by employing an instrumental variable approach and a Heckman selection model. Finally, we control for the potential relationship banking effect and for the heterogeneity of lender banks. In Appendix B, we report some additional results. First, we verify whether firms' scores in each CSR pillar have a different impact on loan spreads. Second, we test whether the impact of CSR ratings is moderated by firm credit quality. Finally, we verify whether the impact of CSR rating changes in crisis times.

#### **7. Discussion**

Our results outline a complex picture of the relationship between CSR ratings and loan spreads. The previous studies offer mixed findings on the link between CSP and credit risk in the European context. Unlike existing studies investigating the European context, we explicitly examine potential nonlinearities in the relationship between CSP and the cost of debt, and we find evidence of a significant lack of homogeneity within the European Union. Bae et al. [15] find a U-shaped relationship between CSR investments and debt financing costs in a sample of syndicated loans issued by U.S. firms. We confirm Bae et al. [15] results just for low ESG-rated European countries, whereas in high ESG-rated countries, we find evidence consistent with the hypothesis that a firm's loan spread declines as the firm's CSR rating improves. The existing literature (among others, Utz [43] and Stellner et al. [5]) has clearly proved that the link between CSP and debt financing costs is highly country specific. Our results offer further confirmation to the hypothesis that, for what concerns the link between CSP and the cost of debt, the relationships observed in the US context do not necessarily hold in the European context. Our findings are undocumented in the existing literature concerning the debt financing cost of European firms. Evidence of nonlinearities in the relationship between CSP and CFP is provided also by Utz [43], who finds a U-shaped relation between CSP and idiosyncratic risk in the United States and in the Asia-Pacific region. For what concerns the European context, Utz [43] finds that European firms have their maximum idiosyncratic risk at a very low level of the CSP score. Beyond that level, a higher score always reduces the idiosyncratic risk. However, Utz [43] treats European firms as a homogeneous sample and does not control for any country specific variable.

In order to compare our results with the existing literature, below we focus our attention on studies whose sample include European firms. Menz [4] shows that companies with better CSP face, respectively, higher spreads for their corporate bonds and a higher cost of debt. The observed differences in the results may be due to several factors (e.g., differences in sample and/or methodology). We emphasize that, first, the adopted CSP indicator may have a limited explanatory power, as acknowledged by Menz himself. Second, the time period investigated by Menz ends in 2007. Given that social raters have gained increasing attention in recent years, the relevance of CSR ratings could have considerably changed over the last years. Third, Menz studies a sample including credit-rated firms only, while in our sample only 40% of companies have a credit rating. Fourth, the European corporate bond market suffers from significant liquidity problems. As far as we know, the author does not control for variations in the liquidity premium paid by corporate bonds in his sample. Fifth, the author does not take into account the impact that the national institutional context exerts on the domestic financial markets. In contrast, we control for the different models of capitalism. In this way, we are able to account for the different institutional context in each cluster, and we find that companies' credit risk is affected by the characteristics of the domestic financial market. Finally, unlike us, Menz does not investigate whether the country ESG performance moderates the CSP–credit risk link.

Our results are more in line with those of Stellner et al. [5]. These authors find some evidence that superior CSP results in lower credit risk in the European corporate bond market, although the statistical significance of their results is rather weak. However, we recall that less than 3% of medium and large European companies have access to the corporate bond market. In contrast, almost all European companies have access to bank financing, and our findings show that banks consider the CSR rating in their loan pricing decisions. In line with our results, Stellner et al. [5] find support to the hypothesis that countries ESG performance moderates the CSP–credit risk relationship and that superior CSP is rewarded with lower bond spreads only if it is recognized by the environment. Compared to Stellner et al. [5], our study offers significant additional insights. Our results provide a more complete picture, as we prove the existence of nonlinearities in the relationship between CSR and loan spreads in low ESG-rated countries, and we are able to highlight the differences between high and low ESG-rated countries in the European context.

Particular attention is required for the findings of Hoepner et al. [6], who examine the syndicated loan market based on a sample that includes borrowers belonging to 28 different countries located in different geographical regions: America (excluding the United States), Asia, Europe, and the United States. They find no conclusive evidence that firm-level sustainability influences the interest rates charged to borrowing firms by banks. The differences between their results and ours may be due to several factors. For example, we note that Hoepner et al. [6] may not take properly into account the impact the national institutional context exerts on loan spreads. Although they include in their models the country's sustainability rating, they control for different national institutional contexts only by means of a dummy that distinguishes developing countries from developed countries. This methodology may not take properly into account the heterogeneity across countries. Country ESG performance is affected by national institutions, but at the same time, these institutions directly affect the cost of the loan. For example, in line with the existing literature, we claim that the legal protection that a country's legal system grants to creditors may significantly affect the cost of bank debt. The legal protection of creditors' rights is not at all considered in a country sustainability rating. In contrast, we take into account the previous variable by clustering countries into different groups according to the VoC approach.

Furthermore, unlike Hoepner et al. [6], we present a cross country analysis, splitting our sample into high ESG-rated countries and low ESG-rated countries. It is only through this analysis that we are able to identify the differences between the two groups of countries and adequately grasp the impact of the CSR rating on the cost of bank debt.

Finally, with regard to the loan characteristics, Hoepner et al. [6] only control for the maturity of the loan, omitting other variables relevant to pricing (collateral, covenants, seniority, and loan type). These variables significantly affect loan spreads in our models, in line with the existing literature.

#### **8. Conclusions**

This study examines the impact of CSR ratings on syndicated loans spreads charged to European listed firms. We find that the CSR rating level affects loan spreads, as lower CSR ratings are on average associated with significantly higher spreads. However, the relationship between these two variables is quite complex. Looking at the whole sample, companies in the highest tertile of the CSR rating distribution do not pay significantly lower spreads than companies in the median tertile. A more detailed investigation allowed us to verify that the home country ESG rating sharply affects the relationship between CSR ratings and loan spreads. In summary, first, low CSR ratings levels are generally penalized with higher spreads by lenders. Second, high CSR rating levels lead to lower loan spreads only for companies located in countries with a high sensitivity to ESG issues. Third, in low ESG-rated countries, firms with high or low CSR rating pay higher loan spreads compared to firms with median CSR rating, providing evidence of a U-shaped relationship between CSR ratings and loan spreads. These results are consistent with the overinvestment view beyond an "optimal threshold" of CSR engagement.

Our results also suggest that CSR ratings are second-order determinants of loan spreads, which are taken into account only after "traditional" firm's fundamentals (i.e., accounting data and credit ratings).

Our findings have significant implications for managers, firm's stakeholders, and legislators. Poor CSP, "certified" by low CSR ratings, leads to greater borrowing costs. However, the efforts and the investments needed to gain high CSR ratings are rewarded only if the company operates in contexts that pay attention to CSR-related matters. Knowing the relationship between CSR-related activities and credit spreads helps managers make appropriate strategic investments in CSR activities. In addition to managers, lenders and outside investors can also rely on the CSP–cost of debt link to assess the firm's future credit health. Unfortunately, firms do not automatically benefit from high CSR ratings. The link between CSP and CFP is conditional on other important variables, some of which are beyond the control of firm's managers (e.g., the home country sensitivity to ESG issues).

Overall, our results suggest that CSR ratings could be a particularly useful tool for less informed stakeholders, such as consumers and retail investors who are interested in evaluating and comparing the CSP of different companies. From this perspective, the CSR rating may also improve firm's accountability and allow cross-company comparisons. Third-party external verification provided by specialized rating agencies enhances the reliability of CSR-related activities, inasmuch as it helps to bridge the credibility gap between the company's self-laudatory CSR communication and less informed stakeholders.

For what concern policymakers, our study offers some support for the vision of the European Parliament, mentioned in Section 1. However, the road ahead to support the transition to a more sustainable economy seems still very long. Given the lack of homogeneity detected in the European context, policy makers should be aware that a uniform legislation on CSR matters could have a mixed impact on the financial performance of companies located in different European Union countries. Mandatory investments in CSR do not necessarily create value for all EU companies. At present, low ESG-rated countries seem to value a high commitment in CSR as a luxury that companies cannot afford. At the same time, we doubt that law could enforce investors to reward a firm's CSR engagement. In order to achieve the objectives of the European Commission, apart from legislative measures, we believe it is necessary to promote several initiatives that support a change in the cultural attitude toward CSR and sustainability issues. Unfortunately, this is a challenging and time consuming process.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2071-1050/12/18/7639/s1, Table S1: Variables description. Table S2: Correlation matrix. Table S3: The impact of CSR ratings on loan spreads. Table S4: Variable description. Table S5: The impact of firm's scores in each CSR pillar on loan spreads.

**Author Contributions:** The authors contributed equally to the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors wish to acknowledge many helpful comments from Raffaele Gallo (Bank of Italy). The usual disclaimer applies.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A. Robustness Tests**

In the following sections, we address potential endogeneity issues by employing an instrumental variable approach and a Heckman selection model. Finally, we control for the potential relationship banking effect and for the heterogeneity of lender banks.

#### *Appendix A.1. Instrumental Variable Approach*

To address potential endogeneity and reverse causality issues, we estimate Equation (1) considering the impact of *EW\_Rating* on loan spreads by employing an instrumental variable (IV) approach. To this end, we instrument *EW\_Rating* with *EW\_Rating\_Lag*, which is the lagged CSR rating of the *i*-th firm from 3 years before *t*.

In line with Goss and Roberts [8], we assume that this instrument is valid. We exclude that our instrument is weak, because high CSR rating levels are obtained after several years of CSR efforts and should be more persistent than financial performance indicators. Thus, the current CSR rating level of the borrower should be significantly affected by the lagged CSR rating level. The first-stage F-test statistic is equal to 57.71, which is significantly above the "rule of thumb" threshold of 10. Therefore, the F-test confirms the significance of our instrument.

In addition, we also assume that our instrument is exogenous. The CSR scores assigned to firms 3 years before the loan closing date, in fact, should be unlikely to affect the loan spreads applied in *t*. Consequently, the instrument should not affect loan spreads other than through its correlation with the current CSR rating level of the borrower. Table A1 reports the estimates from the second-stage IV regression (we observe a reduction in observations because not all CSR-rated borrowers also had a CSR rating three years before *t*). The results confirm that the instrumented variable *EW\_Rating* has a significant negative impact on loan spreads. Consequently, these estimates mitigate endogeneity concerns.



Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

#### *Appendix A.2. Heckman Selection Model*

Since firms choose their levels of engagement in CSR activities, self-selection bias may represent a potential concern for our analysis. To address this issue, we adopt a Heckman [65] approach. To this end, we expand our sample by including also loans registered in LPC's DealScan to European listed companies without CSR ratings. Our new dataset consists of 1727 loans granted to 483 companies.

The first-stage selection equation is a probit model where the dependent variable is a dummy equal to 1 if the *i*-th firm has a CSR rating. To identify the selection equation, in line with Goss and Roberts [8], we add *NWC*, *OpInc*, and *RET*, which are, respectively, the ratio of net working capital, operating income, and retained earnings to total assets. In addition, we also include in the selection equation all variables included in the vector *B* (*Borrower Variables*), *Sov\_Rating*, VoC dummies, and year dummies.

We calculate the inverse Mills ratio from the selection equation and we include it (variable *Lambda*) in the loan spread equation (Equation (1)), to control for selection bias. Table A2 shows the results of this test (we observe a loss of 26 observations due to some missing in the time series of *NWC*, *OpInc* and *RET*). The inverse Mills ratio is significant, suggesting potential self-selection effects. However, the coefficient of *EW\_Rating* is not affected by the inclusion of *Lambda,* supporting our findings.


**Table A2.** The Heckman selection model.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

#### *Appendix A.3. Potential Relationship Banking E*ff*ects and Lender Characteristics*

Our findings could be driven by the potential effects of the relationship between banks and borrowers. Since customer relationships generate private information to banks about their clients, previous relationships between banks and their borrowers could lead to lower borrowing costs [58,66].

To address this issue, we consider as "relationship lenders" the arrangers that were at least in one syndicate of a loan granted to the same borrower before the current loan. We focus on arranger banks because they assess borrower quality, negotiate loan contract terms, and only then do they invite and coordinate participant banks [67]. Following previous studies [57,68], we consider each facility multiple times to capture the differences across arrangers if there are multiple arranger banks in the same syndicate.

To control for previous relationship banking effects, we include in our model the variable *Relationship*, which is a dummy variable equal to 1 if the arranger was in a syndicated loan granted to the *i*-th firm prior to the current loan in the investigated period (Following Goss and Roberts [8], in our previous estimates we have indirectly controlled for potential relationship banking effects by including the variable *Loan\_Concentration*. The variable *Relationship* allows us to take directly into account this potential factor, which could significantly affect loan spreads [69]).

In addition, we control for lender and syndicate characteristics by including *Share*, which indicates the share of the loan to the *i*-th firm held by each arranger; and *NumLenders*, which is the number of lenders in the syndicate. Finally, we include bank fixed effects. Table A3 shows the results. As expected, previous relationships with the same lenders lead to lower loan spreads. However, the coefficient of *Relationship* is rather small. This is not surprising, because the literature finds that the previous relationship produces greater benefits for unlisted companies than for listed ones [58,66]. It is worth noting that our sample includes only listed companies.

We find that the share of the loan held by arrangers is positively related to loan spreads. In fact, the loan share concentration is generally positively related to the borrowers' risk, because arrangers frequently hold a greater stake in the loan if the borrower requires more intense monitoring [66,70].

More lenders in the syndicate are associated with greater loan spreads. We underline that in our sample, the number of lenders is positively related to the number of foreign banks in the syndicate. Therefore, the positive coefficient of *NumLenders* may be due to the expansion of the set of creditors to less-informed investors, such as foreign banks, which, consistent with Sufi [71], may require a greater spread to participate in the loan syndicate.

Finally, since the coefficient of *EW\_Rating* remains significant and negative and our main findings remain unchanged, we can confirm that our results hold also controlling for relationship banking effects and other lender and syndicate characteristics.


**Table A3.** The impact of CSR ratings on loan spreads taking into account lender bank characteristics.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

#### **Appendix B. Additional Results**

In the following sections, we report additional results. First, we verify whether firms' scores in each CSR pillar have a different impact on loan spreads. Second, we test whether the impact of CSR ratings is moderated by firm credit quality. Finally, we verify whether the impact of CSR ratings changes in crisis times.

Table S4 in Supplementary Materials shows the complete list of variables used in Appendix B. Table A4 provides summary statistics of the variables used in Appendix B.


**Table A4.** Summary statistics.

Notes: The sample consists of 1101 loans granted to listed non-financial firms established in EU member states during the 2006–2015 period.

#### *Appendix B.1. The Impact of Di*ff*erent CSR Pillars*

To verify whether firms' scores in each CSR pillar have a different impact on loan spreads, we introduce the following variables: *Ec\_Score*, *Soc\_Score*, *Env\_Score*, and *Gov\_Score*, which represent the ASSET4 scores in each pillar (economic, social, environmental, and governance) of the *i*-th firm in the year preceding t. Then, we replace *EW\_Rating* by alternatively inserting in Equation (1) each of the previous variables. We get four different models whose results are presented in Table A5 (the complete results are reported in Supplementary Materials in Table S5). We observe that all scores are significantly and negatively related to loan spreads, suggesting that better scores lead to lower firms' borrowing costs. An increase in each CSR dimension of 10 scores reduces the average loan spread applied to borrowers by 2.2% for the economic pillar, 4.2% for the social pillar, 4.1% for the environmental pillar, and 1.4% for the governance pillar, respectively. Therefore, banks also appear to positively evaluate firms' efforts in each CSR dimension.

The analysis of the individual scores offers further insights. *Ec\_Score* is highly significant, but its coefficient is about half compared to the *Soc\_Score* and *Env\_Score* coefficients. We believe that the reduced impact of the economic score on the loan spread can be explained by considering that lenders are able to obtain most of the economic data using traditional accounting information derived from the annual report. Furthermore, we highlight that, in line with the existing literature, in the regression analysis we include several variables (e.g., *ROS*, *IE\_Revenue*, etc.) that are able to capture some relevant economic information affecting syndicated loan spreads. However, the high significance of the *Ec\_Score* coefficient shows that the information provided by the CSR economic score is appreciated by the lenders, and it is perceived as supplementary to other economic information of strict accounting derivation. The economic pillar combines key performance indicators (KPIs) based on wider economic information. For example, *Ec\_Score* includes measures about a company's capacity to improve its margins by the use of advanced cost/risk management techniques, or a company's management commitment and effectiveness toward generating sustainable and long-term revenue growth, while maintaining a loyal client base through satisfaction, programs, and avoiding anti-competitive behaviors and price fixing.


**Table A5.** The impact of firm's scores in each CSR pillar on loan spreads.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

Focusing on the *Soc\_Score* and *Env\_Score*, we note that the two coefficients present a high significance and magnitude. These results can be explained by considering that the social and environmental pillars offer important information to lenders not captured by traditional financial information. Furthermore, the performance levels in the environmental and social dimensions signal the company's commitment to CSR dimensions that may originate important levels of risk. Several corporate scandals (e.g., Bayer, Volkswagen, etc.) have repeatedly shown that bad performances in the environmental and social dimensions can expose companies to significant losses and negative market assessments. It is not surprising that the attention and sensitivity of the lenders is mainly focused on these two pillars.

The reduced significance and the magnitude of the *Gov\_Score* coefficient do not surprise. This result could be explained considering that lenders have other sources available to derive information on governance, especially for large companies such as those included in our sample. However, the information provided by the governance score seems not to be superfluous. Lenders appreciate the information provided by the governance score for a broader assessment of the risk levels of their borrowers. Our result can be explained considering that the corporate governance pillar includes several measures about a company's systems and processes, which ensure that its board members and executives act in the best interests of its long term shareholders. It reflects a company's capacity, through its use of best management practices, to direct and control its rights and responsibilities through the creation of incentives, as well as checks and balances in order to generate long term shareholder value.

Finally, we note that in all estimated models, our control variables maintain the signs and the statistical significance discussed in the main text.

#### *Appendix B.2. CSR Ratings and Creditworthiness*

Previous studies found that the creditworthiness of the borrower could significantly moderate the link between CSR ratings and loan spreads [8]. We argue that, given the same CSR rating, low credit quality firms pay greater loan spreads than high credit quality firms. We distinguish two cases that lead to the same conclusion. First, if the CSR rating is low, having a low creditworthiness has a multiplicative effect on the risk of the borrower. In this case, the borrower is exposed to significant CSR-related risks in addition to "traditional" financial risks. Second, if the CSR rating is high, CSR investments made by low quality firms are not rewarded by lenders. In fact, since less creditworthy firms have fewer available resources than safer ones, proactive engagement in CSR and greater discretionary investments may be perceived by lenders as a costly diversion of scarce resources.

To verify whether the impact of CSR ratings is moderated by firm credit quality, we identify less creditworthy firms by alternatively adopting three sets of variables. First, we rely on the z-score, an accounting measure that indicates the probability of firms' bankruptcy. Z-scores are calculated using firms' quarterly data over the past 3 years following Santos and Winton [72]:

$$Z = \frac{1}{S\_r} \left[ \frac{1}{n} \sum\_{j=1}^n \frac{2\tilde{\pi}}{A\_j + A\_{j-1}} + \frac{1}{n} \sum\_{j=1}^n \frac{E\_j + E\_{j-1}}{A\_j + A\_{j-1}} \right]$$

where π is the firm's profits, *A* is its assets, *E* is its equity, and *S<sup>r</sup>* is the estimated standard deviation of *r*, the firm's return on assets.

We introduce in Equation (1) two dummy variables: *HighZscore* and *LowZscore*. These variables are equal to 1 if the z-score of the *i*-th firm in the year preceding *t* is, respectively, in the highest or in the lowest tertile of the empirical distribution. We interact *EW\_Rating* with *HighZscore* and *LowZscore,* respectively. We expect to find that higher CSR ratings lead to lower spreads mainly for safer firms (higher z-scores).

Second, we rely on the dummy variable *Secured*. Empirical evidence has demonstrated that lenders demand security mainly from low-quality borrowers [73]. Thus, we add in Equation (1) the interaction variable *EW\_Rating*·*Secured*. We expect that, given the same CSR rating, secured loans are charged with higher spreads.

Third, we identify riskier firms by relying on their size. Holding all else equal, smaller firms are generally considered riskier than larger ones, because they are less transparent and more financially constrained. Thus, we replace the variable *Size* in Equation (1) with *Small*, which is a dummy variable equal to 1 if the *i*-th firm's total assets in the year preceding *t* are lower than the tenth percentile of the sample. We underline that, since small firms do not have generally access to the syndicated loan market, the majority of firms in our sample are large. Therefore, we have adopted a low threshold to identify smaller firms. Moreover, we include in our model an interaction between *EW\_Rating* and *Small.* A positive and significant coefficient of the interaction variable would imply that, given the same *EW\_Rating*, low quality borrowers pay higher spreads than high quality borrowers.

We highlight that the interaction between *EW\_Rating* and *Small* allows us to capture the effect of the firm's size on the impact of CSP on loan spread. In particular, we are able to verify whether, all other things being equal, the impact of CSP on loan spread changes for small companies compared to large ones.

Column (1) of Table A6 shows the results of Equation (1) estimated by interacting *EW\_Rating* with *HighZscore* and *LowZscore* (We observe a reduction in observations due to data availability). Our control group consists of firms with median *Zscore*. We observe that *EW\_Rating*·*HighZscore* has a negative impact on loan spreads, while the interaction variable *EW\_Rating*·*LowZscore* is positively correlated with *LnSpread*. This implies that the impact of CSR ratings on loan spreads significantly depend on the probability of default of the borrower. Riskier firms (lower z-scores) always pay higher loan spreads than the control group, given the same CSR rating level. In contrast, safer firms (higher z-score) pay lower loan spreads compared to the control group. Therefore, our findings suggest that banks reward greater CSR efforts mainly when the credit risk of the borrower is low, whereas CSR investments of riskier firms may be perceived as a costly diversion of scarce resources.

Column (2) of Table A6 shows the results of Equation (1) estimated by interacting *EW\_Rating* with *Secured*. The interaction term *EW\_Rating*·*Secured* shows a positive impact on loan spreads and counteracts the negative coefficient of *EW\_Rating*. An increase in the CSR rating of 10 scores reduces by about 5% the average spread on secured loans and by about 2.1% that on unsecured loans. These results confirm that the negative relationship between CSR ratings and loan spreads is weaker for riskier firms (i.e., those that receive secured loans).


**Table A6.** CSR ratings and the firm's creditworthiness.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

Finally, in column (3), we report the results of Equation (1) estimated by replacing *Size* with *Small* and by adding the interaction *EW\_Rating*·*Small*. We observe that the coefficient of the interaction variable is positive and significant. An increase in the CSR rating of 10 scores reduces by about 6.6% the average loan spread for larger firms and by about 2.1% that for smaller ones. Therefore, the firm's size moderates the relationship between CSR ratings and loan spreads. This result suggests that, given the same CSR rating, smaller firms pay higher loan spreads.

Overall, these findings suggest that the benefits of better CSR ratings are lower for riskier firms. These results may be interpreted as an additional confirmation that CSR ratings are second-order determinants of loan spreads, which are taken into account only after traditional financial factors.

#### *Appendix B.3. Crisis*

The time frame considered in our analysis includes two periods of major crises, the great financial crisis and the sovereign debt crisis, which have heavily affected the European economic system.

Existing literature finds evidence of a greater increase in loan spreads for European companies during the global financial crisis and the euro area sovereign debt crisis [74]. At the same time, we expect that, all other things being equal, in crisis periods, lenders are less sensitive to information related to the borrower's CSP, and they assign greater importance to the borrower's financial data. If this is true, we should observe a lower impact of the CSR rating on loan spreads in crisis periods.

To consider the potential impact of crisis periods on loan spreads we include in Equation (1) the variable *Crisis*, which is a dummy variable equal to 1 in crisis periods of the European economy. To determine crisis periods, we rely on the chronology of turning points for Europe identified by the OECD [75]. Consequently, the periods from February 2008 to June 2009 and from August 2011 to February 2013 are considered crisis periods. In contrast, we consider other periods as non-crisis periods.

Column (1) of Table A7 shows the results of Equation (1) estimated by replacing year dummies with the variable *Crisis*. As expected, the positive sign of *Crisis* suggests that, on average, banks raise loan spreads in crisis times. Moreover, in line with previous results, we find that an increase in the CSR rating of 10 scores reduces by about 3.4% the average loan spread applied to borrowers.


**Table A7.** The impact of CSR ratings on loan spreads considering crisis periods.

Notes: Robust standard errors are in parentheses. \*\*\*, \*\*, and \* denote significance at the 1%, 5%, and 10% level, respectively.

To verify whether the impact of CSR ratings on loan spreads changes in crisis times, we run the previous regression by adding in our model an interaction variable between *EW\_Rating* and *Crisis*. Column (2) of Table A7 shows these estimates. We observe that the coefficient of *EW\_Rating*·*Crisis* is not significant. However, we note that the significance of *EW\_Rating* remains unchanged. Thus, our results show that crisis periods do not significantly affect the impact of CSR ratings on loan spreads. Lenders continue to positively evaluate borrowers' CSR efforts also in these periods.

In light of these results, we can confirm the validity of our main findings taking into account the potential effects of crisis periods.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Wealthy Private Investors and Socially Responsible Investing: The Influence of Reference Groups**

**David Risi 1,\*, Falko Paetzold 2,3 and Anne Kellers 3,4**


**Abstract:** Sustainable development requires a shift from traditionally invested assets to socially responsible investing (SRI), bringing together financial profits and social welfare. Private high-networth individuals (HNWIs) are critical for this shift as they control nearly half of global wealth. While we know little about HNWIs' investment behavior, reference group theory suggests that their SRI engagement is influenced by their identification with and comparison to reference groups. We thus ask: how do reference groups influence the investment behavior of SRI-oriented HNWIs? To answer this question, we analyzed a unique qualitative data set of 55 semi-structured interviews with SRI-oriented HNWIs and industry experts. Our qualitative research found that, on the one hand, the family serves as a normative reference group that upholds the economic profit motive and directly shapes HNWIs to make financial gains from their investments at the expense of social welfare. On the other hand, fellow SRI-oriented HNWIs serve as a comparative reference group that does not impose any concrete requirements on social welfare performance, indirectly influencing SRI-oriented HNWIs to subordinate social concerns to financial profits. Our scholarly insights contribute to the SRI literature, reference group theory, and practice.

**Keywords:** high-net-worth individuals (HNWIs); qualitative research; reference group theory; socially responsible investing (SRI)

#### **1. Introduction**

A shift from traditionally invested assets to socially responsible investing (SRI), broadly defined as the integration of environmental, social, and governance (ESG) considerations into investment practices, is a crucial driver of sustainable development [1]. Millionaires and billionaires, i.e., private high-net-worth individuals (HNWIs), hold a vital role in this shift. The United Nations calculated that investments of USD2.5 trillion per year are missing to finance sustainable development [2]. Thereby, the wealthy top 1% of the world's population controls about USD 191.6 trillion as of 2020, nearly half of global wealth [3]. It is crucial to understand the investment behaviors of HNWIs to mobilize this substantial source of capital for sustainable development.

To understand whether private investors engage in SRI, the literature tends to put a higher emphasis on proving the financial profitability of SRI (see [4–6]) than, for example, its positive impact on social welfare [7,8]. However, since SRI brings together financial profits and social welfare, sustainable investing goes well beyond the question of whether or not SRI is more profitable than traditional investing [6,9–13]. Still, many investors are attracted to SRI due to social welfare reasons (e.g., [14–16]). Consequently, the profitability debate around SRI only partially solves the issue of knowing little about sustainable investors [16,17] and SRI-oriented HNWIs [18,19]. To gain deeper insight into the investment

**Citation:** Risi, D.; Paetzold, F.; Kellers, A. Wealthy Private Investors and Socially Responsible Investing: The Influence of Reference Groups. *Sustainability* **2021**, *13*, 12931. https:// doi.org/10.3390/su132212931

Academic Editors: David Aristei, Manuela Gallo and Olaf Weber

Received: 21 September 2021 Accepted: 17 November 2021 Published: 22 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

behaviors of SRI-oriented HNWIs, we need to understand their individual dealings with both social welfare issues and financial gains in their SRI investments.

A reference group theory perspective suggests that the individual investment behavior of SRI-oriented HNWIs is fundamentally influenced by the groups for which the wealthy private investor has a membership. The reference group theory operates on the principle that individuals always orient themselves to others, as their attitudes, values, and selfappraisals are shaped by their identification with and comparison to reference groups [20]. To establish or maintain individual identification with the reference group, individuals behave, believe, and perceive as the group does [21]. There are two types of reference groups [22–24]. Normative reference groups establish and enforce specific standards which can be considered as norms. Comparative reference groups serve individuals as a point of reference in making evaluations or comparisons without the evaluation of the individual by others in the reference group [23].

Hence, from a reference group theory perspective, SRI-oriented HNWIs' identification with and comparison to a respective reference group significantly influences whether, how, and to what extent they bring together financial profits and social welfare in their investments. Thus, while the influence of normative and comparative reference groups is central to our understanding of the investment behavior of SRI-oriented HNWIs, previous research has not yet addressed this issue. Consequently, our knowledge of HNWIs committed to SRI remains underdeveloped. The main objective of our study is to develop this knowledge, and we thus ask: how do reference groups influence the investment behavior of SRI-oriented HNWIs?

To answer this question, we adopt a qualitative research strategy. Such a strategy is advantageous for developing our knowledge of the investment behavior of SRI-oriented HNWIs because qualitative research supports the generation of novel insights "at a level of detail and nuance that can be difficult or impossible to achieve using only quantitative methods" [25] (p. 637). We conducted semi-structured interviews with 42 SRI-oriented HNWIs and 13 experts who consult with them and closely monitor the SRI market. Based on our analysis of this unique empirical data, we develop a framework to explain how different reference groups influence the investment behavior of SRI-oriented HNWIs. Our framework indicates that, on the one hand, the family serves as a normative reference group that holds up economic profit striving and directly influences HNWIs towards generating financial profits in their investments at the expense of social welfare considerations. On the other hand, fellow SRI-oriented HNWIs serve as a comparative reference group that places little emphasis on accountability for social issues and indirectly influences SRI-oriented HNWIs to subordinate social welfare issues to financial gain.

Our research makes two contributions to the literature. First, we add to SRI research by providing insights into the hitherto little-researched SRI engagement of wealthy private investors (e.g., [16]). Our framework explains that SRI-oriented HNWIs prioritize financial gains at the expense of social welfare because they are encouraged by reference groups to use their wealth to achieve economic profits, even though they already have immense wealth. Second, we contribute to reference group theory, which suggests different reference groups based on differentiating between a normative and a comparative function of a reference group (e.g., [23]). We show that normative and comparative reference groups can coexist but that the normative reference group suppresses the comparative reference group in conflict. This finding implies different spheres of influence of normative and comparative reference groups.

We proceed by presenting existing SRI research on HNWIs and, on this basis, problematizing the lack of knowledge on the influence of reference groups on the investment behavior of SRI-oriented HNWIs. We then outline our research context and method and present the results of our study. On this basis, we develop a framework of how reference groups influence the investment behavior of SRI-oriented HNWIs. We finish by discussing the implications for the literature, some practical implications, the limitations of our study, avenues for further research, and a conclusion.

#### **2. Theoretical Background**

#### *2.1. Socially Responsible Investing (SRI) and High-Net-Worth Individuals (HNWIs)*

Socially responsible investing (SRI) integrates environmental, social, and governance (ESG) issues into investment practice and closely links to sustainable development [26,27]. The peculiarity of SRI, especially compared to traditional investing, is that it combines two different and potentially conflicting logics: while the market logic has the primary characteristic of the pursuit of financial profit, the social welfare logic is grounded on communitarianism, altruism, the fulfillment of social needs, and the solving of social misery (see [6]). Regarding the segment of private SRI-oriented investors, some studies address their characteristics, motivations, and barriers and provide comparisons with non-SRI investors (e.g., [28–32]). Among private investors, those with discretionary investable assets of more than USD 1 million, defined as high-net-worth individuals (HNWIs), are of particular interest [33]. While as of 2020, HNWIs represent 1.1% of the world's population, they hold 46% of global household wealth [3] and can thus contribute significantly to the growth of SRI. HNWIs tend to be interested in incorporating SRI aspects, such as climate change, into their investment decisions, as they "are typically long-term investors whose aim is to preserve capital for the next generations to come" [33] (p. 7). Moreover, HNWIs are in the position where they can invest along with their personal interest because they "have access to investments that are normally closed to smaller retail investors, and the freedom to move funds quickly without having to perform the extensive due diligence required by institutional investors" [33] (p. 7).

To understand whether private investors engage in SRI, the academic literature puts a higher emphasis on the ability to prove the financial profitability of SRI (see [4–6]) than, for example, its positive impact on social welfare [7,8]. However, since SRI brings together financial profits and social welfare [6], sustainable investing goes well beyond the question of whether or not SRI is more profitable than conventional investing, as, evidently, "there are more nuanced issues at stake than just profits" [9] (p. 360) (see also [10–12]). Similarly, Revelli [34] (p. 711) critically notes that in the course of the efforts around the mainstreaming of SRI, "the original goal of 'making good'" has transformed "into a quest for profitability".

Addressing profitability can only help us understand to a limited extent whether investors are committed to SRI, as many investors are attracted to SRI due to altruistic motives [15,16]. For example, a study by Barreda-Tarrazona, Matallin-Saez, and Balaguer-Franch [14] shows that although diversification and return are essential drivers of SRI investment, private investors, who embrace SRI, tend to invest in SRI funds even when the return differential is negative. In their review of the SRI literature, Renneboog, Ter Horst, and Zhang [35] conclude that prior research suggests that SRI investors are willing to accept suboptimal financial profits to contribute to social welfare. The latter research supports the rising voices of scholars questioning the "business case" justification and associated profit maximization arguments for socially responsible business practices (e.g., [36,37]) and SRI (e.g., [6,34]). Juravle and Lewis [38] confirm this by showing that investors often do not engage in SRI because of cognitive patterns and normative belief systems. They note that even experienced investors are susceptible, for example, to herd behavior or fads and are guided in their investment behavior by the belief of the incompatibility of financial profit and social welfare.

Consequently, the profitability debate around SRI can only partially solve the circumstance of still knowing little about sustainable investors [16,17] and SRI-oriented HNWIs [18,19]. In contrast, a deeper insight into the investment behavior of SRI-oriented wealthy private investors requires that we go beyond this very debate and understand how HNWIs deal with social welfare issues and financial profits in their SRI investments. The point here is to consider that the individual investment behavior of SRI-oriented HNWIs is always shaped by the group in which the wealthy private investor has a membership. Therefore, we introduce the reference group theory, which points out that individuals

orient themselves to others, so-called reference groups, and thus individual thinking and acting are fundamentally shaped by others.

#### *2.2. Reference Group Theory Perspective on the Investment Behavior of SRI-Oriented HNWIs*

Generally, a reference group has been defined "as a group, collectivity, or person taken into account by an actor and used in such a manner that he identifies himself and uses the group, collectivity, or person as a basis for self-evaluation and as a source of his personal values and goals" [39] (p. 68). As this definition suggests, reference group theory builds on the assumption that human beings desire the feeling of oneness with groups [21]. Such non-formalized memberships give people the confidence that the appropriate strategies to manage one's life are befitting and valid [20]. To obtain this group identity, one needs to behave, believe, and perceive as the group does [20,21,40,41] and socialize oneself to what one perceives to be the group's norms [42]. Consequently, an individual's attitudes, values, and self-appraisals are influenced by the identification with and comparison to reference groups [20]. This includes articulating and reasoning things important to oneself so that others will accept these explanations of what constitutes important [20]. Hence, the reference group influences the behavior of individuals due to anticipation of the responses of the group [43].

Reference group theory distinguishes between normative and comparative reference groups [22–24]. Normative reference groups are groups where individuals are motivated to establish or maintain acceptance. To reach that goal, individuals keep their attitudes in conformity with what they perceive to be the consensus of opinions (norms) among their reference group [20,23]. Here, the group establishes and enforces specific standards which can be considered as norms. Consequently, the normative function of a reference group is that it provides individuals with a basis for forming goals and values and expects them to comply with the goals and values of their reference groups [39]. Values are normative beliefs that guide human actions, as they specify "the things that are worth having, doing, and being" [44] (p. 356; see also [45]). Values are particularly central in normative contexts when, as in the case of SRI, it is a matter of conceptualizing the respective possibilities and limits in reconciling economic and social aspects [46].

On the other hand, comparative reference groups serve individuals as a point of reference in making evaluations or comparisons [23]. In a comparative reference group, the evaluations of the individual by others in the reference group are irrelevant. The group serves as a standard or checkpoint that the individual uses to make judgments [23]. The comparative function of a reference group thereby provides a frame of reference that an individual uses for self-evaluation, thus resulting in either a satisfactory or unsatisfactory view of oneself [39]. From a reference theory perspective, SRI-oriented HNWIs seek nonformalized membership in groups to gain the confidence that their investments are befitting and valid. In doing so, SRI-oriented HNWIs align their attitudes and behaviors toward investment with what they think the respective reference group expects of them. For example, in the case of other wealthy private investors, we would assume that HNWIs make economic success observable through their investment activities and behavior to maintain "social prestige" or "social status" within the group [47–49]. Financial profit would signal that the individual HNWI is adapting to what she or he thinks is necessary for membership in the reference group (in this case, other HNWIs).

Also, HNWIs regularly discuss their investment decisions with family members [18], suggesting that this group may serve as a basis for HNWIs' self-assessment and personal values and goals. At the same time, SRI-oriented HNWIs are, of course, also influenced by other like-minded HNWIs. In this reference group, one would assume that members hold up and demand not only financial profit but at least equal claims regarding social welfare and expect that group members meet these standards. Hence, by contributing to social welfare through investments, an individual HNWI portrays that she or he behaves, believes, and perceives as the group of other SRI-oriented HNWIs does.

Unfortunately, there is no research on how reference groups influence HNWIs' SRI engagement, even though the literature suggests that they would fundamentally influence how SRI-oriented HNWIs deal with social welfare issues and financial gains in their investments. Hence, our knowledge of the investment behavior of HNWIs committed to SRI remains limited, and our research correspondingly asks the following question: how do reference groups influence the investment behavior of SRI-oriented HNWIs?

#### **3. Methods**

We apply a qualitative inductive research design to gain detailed insights into how reference groups influence the investment behavior of SRI-oriented HNWIs. Because of the nascent nature of theory in the context of SRI-oriented HNWIs (see, e.g., [18]), it is necessary to take a qualitative approach that ensures a "methodological fit" with our research endeavor [50]. For example, Bettis et al. [25] (p. 637) have indicated qualitative approaches as essential tools to generate new insights that document phenomena "at a level of detail and nuance that can be difficult or impossible to achieve using only quantitative methods" (see also, [51]).

#### *3.1. Sampling Strategy and Data Collection*

We use a purposeful sampling strategy aimed at gathering information-rich data sources "from which one can learn a great deal about issues of central importance to the purpose of the inquiry" and that provide "insights and in-depth understanding rather than empirical generalizations" [52] (p. 230). In contrast to approaches such as random sampling, purposeful sampling implies that the selection of data sources runs parallel to the data collection [53]. Simultaneously selecting and collecting the data increases the possibility of generating novel concepts and identifying theoretical relationships with information that either substantiates them or provides divergent examples [54].

We collected our data in the form of 55 semi-structured interviews with HNWIs and industry experts between 2015 and 2019 with the help of wealth owner networks in Europe and the United States. These interviews lasted, on average, 30 min, were recorded, and were fully transcribed. We interviewed 42 SRI-oriented HNWIs with different cultural backgrounds and sources of wealth creation (see Table 1). In the course of these interviews, we asked them about the role of wealth in society, their thoughts around considering ESG criteria in their investments, and their assessment of the importance of SRI for sustainable development. Our questions also addressed their understanding of SRI, the barriers they face, the values and beliefs they hold, and their expectations. Expectations included broader ideas such as overall visions and hopes for the SRI market and particular aspects such as financial return and social welfare contribution regarding their own SRI engagement.

**Table 1.** Overview of informants and some background information.



**Table 1.** *Cont.*

We adopted a range of measures to enhance the reliability of our interview data. We posed "courtroom questions" [55] (p. 41) by asking SRI-oriented HNWIs the same questions to reduce self-reported biases. This technique helps to avoid speculation and enhances the reliability of the informants' responses. As is standard in qualitative research (e.g., [56]), we granted anonymity to all informants to elicit candid responses [55]. Furthermore, we interviewed 13 experts who regularly consult with SRI-oriented HNWIs and closely monitor the SRI market, including advisors, managers, and researchers. This data was relevant for triangulating the interview data gained from the wealthy private investors.

Table 1 provides an overview of all our informants. The table typifies the informants into wealth owners and industry experts, with the latter further subdivided into advisors, managers, and researchers. In addition, the table includes information on each interviewee's age, gender, nationality, country of residence, profession, approximate wealth, and highest academic degree.

#### *3.2. Data Analysis*

We used grounded theorizing and, more specifically, the "Gioia methodology" [57] to analyze our interview data. The Gioia methodology helps analyze interview data in the context of individuals concerned with social and environmental issues in a business context (see, e.g., [58]). This methodology is tailored to qualitative inductive inquiry and comprises three levels of abstraction [57].

The first-order analysis is about processing the raw interview data to identify a primary set of codes. We classified those codes into different groups of descriptions that

our informants provided. This initial assessment provided insights into what SRI-oriented HNWIs consider the prevalent problems that modern societies face and the potential ways to solve them, from political actions to philanthropy and sustainable investing. We have learned what role private wealth plays in this discussion, what opportunities wealthy persons have for adding to social welfare, and what responsibility they ascribe to themselves in this context. Moreover, we obtained preliminary knowledge of what role fellow HNWIs and their family members play in their SRI engagement. The result of this initial stage of analysis were several first-order category codes.

We then engaged in a second-order analysis. We analyzed additional data and studied the literature to incrementally move from the first-order insights toward more theoretical second-order themes. We continuously iterated back and forth between data and literature and gradually developed theory [59]. At this stage, we particularly noticed that SRIoriented HNWIs see wealth as a cause and solution for societal problems and feel personally responsible to society. Furthermore, we learned how the latter use SRI to make financial profits, what their families expect from them, and how SRI-oriented HNWIs try to meet these exact expectations. Moreover, we realized the importance of their peers with whom they share the same values, goals, and visions. The importance of like-minded wealthy private investors and families prompted us to review the literature on reference theory in-depth, stimulating a related oscillation between theory and empirical data. The result of this analysis was a set of second-order themes.

We processed additional data to identify the interaction between key constructs on the highest level of analysis leading to aggregate dimensions. More specifically, we categorized raw data, linked first-order categories to second-order themes, and aggregated them into third-order dimensions. The result was five aggregate dimensions: first, using one's own fortune to promote social welfare; second, using one's own fortune to generate financial profits; third, one's family sets profit-oriented norms; fourth, proving one's profit to conform with family norms; and fifth, other SRI-oriented HNWIs provide confirmation.

Throughout the data analysis, we ensured intercoder reliability. To this aim, we used the data analysis software NVivo. This software helps organize large amounts of qualitative data and provides the basis for performing data analysis in a team. The authors held regular meetings to cross-check the coding and ensure the development of the same understanding of the emerging categories, moving from open coding over more theoretical categories to aggregate dimensions. Figure 1 shows our data structure and, thereby, provides an overview of the three levels of abstraction in line with the Gioia methodology. In this vein, the figure depicts our inductive reasoning process from empirical raw data in the form of first-order categories over second-order themes to more abstract theoretical categories in the form of aggregate dimensions.

In the following findings section, and according to conventions in qualitative research (e.g., [60]), we offer power quotes throughout the text and, per subsection, provide additional interview data supporting our empirical analysis in Tables 2–11.

**Figure 1.** Data structure.

#### **4. Findings**

We structure the empirical results as follows: first, we outline how HNWIs use their own fortunes to promote social welfare. Second, we show that they use their fortunes to generate financial profits. Third, we depict how the family sets profit-oriented norms. Fourth, we demonstrate that SRI-oriented HNWIs engage in proving profit to conform with family norms. Finally, we present how other SRI-oriented HNWIs provide confirmation.

#### *4.1. Using Own Fortune to Promote Social Welfare*

When asked about their motives for SRI, HNWIs often pointed out that they strive to use their fortune to promote social welfare. In the following, we will discuss two aspects of our data supporting this insight.

*Wealth as a cause and solution for societal problems*. Wealth has an essential role in society in that it functions equally as a cause of and solution to societal problems such as inequality. Firstly, many HNWIs describe wealth as the cause by pointing out that *wealth concentration is a societal problem*. One informant (HNWI 12), for example, problematizes wealth concentration by arguing that "wealth distribution is definitely something that I adhere to" in my investment decisions because "I just feel like opportunities are a little bit skewed at this point." Further, the wealth owner problematizes wealth concentration by contrasting it with an equal society that is much more beneficial for all involved, as it ensures equal opportunities, i.e., "a much more balanced society is extremely beneficial for all".

Secondly, HNWIs emphasize that ample financial resources may serve to tackle social problems. One wealth owner (HNWI 16) illustrates *wealth as an important tool for social welfare promotion* by the example of an investment strategy aimed at combating climate change and all its resulting societal consequences. According to this informant, investing wealth through this strategy serves "to bend emissions and create opportunities to generate land that we are able to move back towards a healthy planet." In this regard, the strategy goes far beyond combating climate change by securing that "people are going to be less hungry, be better fed, have better sanitation, and all those things that potentially come with making better use of the resources we have".

In sum, our empirical analysis of the interview data shows that HNWIs see wealth as both a cause of and an opportunity to solve societal problems. On the one hand, HNWIs localize the concentration of wealth as the cause of the unfair distribution of opportunities in society; on the other hand, they describe wealth as the central means of solving current social problems, such as the unfair distribution of resources. In Table 2, we provide further evidence of wealth as a cause and solution for societal problems.

**Table 2.** Wealth as a cause and solution for societal problems.


*Perception of personal responsibility towards society*. The interviewed HNWIs deal in detail with the connection between wealth and the potential responsibility that comes with it and how this very connection affects them personally. Firstly, HNWIs often mentioned the issue of being *guilty of being rich*. For example, after being asked by the interviewer about the fairness debate around inherited wealth and first-generation wealth and how the respective generation and the family as a whole deal with this debate, one informant (HNWI 2) responded that "we know [about the fairness debate around inherited wealth], and it's something that my mom, I think, makes a big effort of reminding us about." Furthermore, the informant explicitly points out the feelings of guilt that come along with being wealthy: "but yes, I do think there's a big element of unfairness there".

Secondly, our data on HNWIs suggest that wealth obliges one to make a positive social contribution. The interviewees clearly express a personal desire to do something about the inequality in today's world and the lack of social mobility. This includes straightforward measures such as the intention to redistribute financial resources but also to use one's own capital to promote projects that increase social mobility. One wealth owner (HNWI 25) clarifies this further by pointing out that "there's this fundamental discomfort with the inequality that exists in the world" and that driving the investment of wealth "at the portfolio level but also the deal level is this sense of how can we create more equality in the world".

To summarize, the interviewed HNWIs see themselves, primarily because of their wealth, as bearing a personal responsibility to society. This sense of personal responsibility is based both on feelings of guilt, which originate from their own wealth, and on the conviction that wealth obliges one to solve social problems such as the increasing inequality

between the rich and the poor. In Table 3, we provide further evidence of the perception of personal responsibility towards society.

**Table 3.** Perception of personal responsibility.


#### *4.2. Using Own Fortune to Generate Financial Profits*

The interviewed SRI-oriented HNWIs expressed that they aim to use their own fortune for generating financial profits, as evidenced by the profit orientation of their sustainable investment activities. We found two aspects supporting this insight that we will detail in the following.

*Financial return is essential*. HNWIs generally regard SRI as a financial instrument that not only has a positive social impact but also generates an economic return. Firstly, this circumstance is shown by the aspect that *SRI needs to pay off*. One wealth owner (HNWI 34) illustrates the importance of making money with SRI by the example of impact investing, which can be understood as a synonym of SRI. This informant notes that people "confuse it [impact investing] with philanthropy" while instead "impact investing is about making a positive impact and make a lot of money".

Secondly, HNWIs often consider their sustainable investing activities as a way of making a financial profit. Hence, wealthy sustainable investors see *SRI as a tool to make some profit*. For example, the following informant (HNWI 1) clarifies the importance of earning money as follows: "The argument is that we don't want to lose money [with SRI]. We don't want this to be an expense. We want to earn money, make investments that are profitable".

In conclusion, our analysis indicates that the interviewed HNWIs conceive SRI as an investment vehicle to contribute to society and generate financial profits. In each case, financial gain is emphasized, for example, when HNWIs point out that SRI should help "make a lot of money" and serve as a tool to generate a financial surplus. In Table 4, we provide further evidence that financial return is essential.

*Profitability to enable the adoption of SRI***.** Profitability has often been expressed under the umbrella of building the field of sustainable investing. Many wealthy private investors mention the need to prove the established idea that SRI should be as equally profitable as traditional investments. This is, firstly, because HNWIs suggest that *profit serves as a primary motive*. One wealth owner (HNWI 31) points out that "the thesis of impact investing is that you can achieve the same returns." Moreover, the informant states that the confirmation of this thesis is critical for whether investors go into impact investing at all: "at the performance of portfolios, there's very little evidence. ( . . . ) If you say that to people, they'll be like, 'hell no, I'm not putting that money into impact'".

Secondly, the interviewed SRI-oriented HNWIs consider *profitability as a compelling argument* to encourage the adoption of sustainable investment practices by third parties. One interviewed HNWI (HNWI 32) explains this by the case of convincing the board of their own family office to adopt SRI: "I had to look at it from the perspective of where can I get some wins, where can I get the leverage going. And it's honestly just about proving that we can make market returns or better".

**Table 4.** Financial return is essential.


In sum, the analysis of the interview data suggests that HNWIs consider the financial profitability of SRI relevant for establishing the field of sustainable investing and for promoting its adoption among wealthy private investors in particular. This insight is grounded on the circumstances that profit motives dominate the investment behavior of HNWIs and that profitability is the most compelling argument for adopting SRI or not. In Table 5, we provide further evidence of profitability to enable the adoption of SRI.

#### *4.3. Family Sets Profit-Oriented Norms*

Families and their members who surround the HNWIs set profit-oriented norms that the wealthy sustainable investors interviewed perceive as standards and expectations they must adhere to. Below, we detail two aspects of the insight that families demand financial profit and claim this demand toward SRI-oriented HNWIs.

*Family upholds the value of wealth preservation and skepticism against SRI*. HN-WIs repeatedly mention the relevance of their family members for their investments. Firstly, their *family upholds the value of wealth preservation* that is an essential guideline for them. Our data suggest, at least in the context of investing, that wealth preservation is the most prominent value in wealthy families. For example, in response to whether there are any particular values or principles regarding financial investments that the HNWI has adopted from their own family, the informant (HNWI 15) mentions values related to "wealth preservation" that many wealthy families have to "set up expectations for family members in order to access funds".

Secondly, the interviewed HNWIs repeatedly point out that *family members are skeptical towards SRI*. Families are often unfamiliar with the underlying idea of SRI, of combining financial investment with a positive social and environmental contribution, and therefore cannot imagine how this would work. One wealth owner (HNWI 8) further explicates this skepticism by "an added level of skepticism that the family office brings whenever we put forth something with the knowledge that it is impact." This informant locates this

skepticism in the technical terms and expressions associated with impact investing, as "they (family office members) themselves put an added level of skepticism on the investments we put forward because of the impact investment terminology".

**Table 5.** Profitability to enable the adoption of SRI.


In conclusion, our informants emphasize that their families uphold the value of wealth preservation and skepticism against SRI. On the one hand, such wealth preservation provides the interviewed HNWIs with a basis for their value formation and presents a critical normative framework against which they align their investment behavior. On the other hand, families are skeptical about SRI and the associated sustainable investment behaviors because, according to the HNWIs interviewed, their family members are often unacquainted with SRI. In Table 6, we provide further evidence of the issue that the family upholds the value of wealth preservation and skepticism against SRI.

*SRI does not fall within the purpose of the family*. HNWIs themselves often face the circumstance that their family does not see the point of linking their financial investments to socially and environmentally positive contributions. Firstly, this circumstance can be explained by the fact that there usually are *family offices without any social welfare mandate*. The following statement by a wealth owner (HNWI 15) illustrates that most family offices lack any mandate for making a positive social or environmental contribution as part of their investment activities: "I know many family offices, and I always ask if they have an impact mandate or something, and a lot of them still don't".

Secondly, SRI-oriented HNWIs mentioned that their families often uphold that they already are engaged in philanthropic activities and therefore do not see any need for SRI. Hence, *family members traditionally consider philanthropy as sufficient*. The extent to which this very attitude can hinder SRI illuminates an informant (HNWI 6) who is appropriately committed to such investments outside the family and its wealth because family members only focus on philanthropy, as explicated by the family foundation: "they (family members) have a very traditional sort of foundation setting. ( . . . ) So the foundation is purely about giving philanthropic capital, not capital but the income generated from it".

To sum up, the analysis points towards the circumstance that HNWIs' families do not see why striving for financial returns should link to a positive societal contribution. This insight reflects the fact that family offices, officially entrusted with managing the family's assets, traditionally do not have a social welfare mandate. Moreover, the circumstance that family members traditionally consider philanthropy to be sufficient, where any economic activity is usually separated from social welfare engagement, supports the insight that SRI does not fall under their families' purpose. In Table 7, we provide further evidence that SRI does not fall within the purpose of the family.

**Table 6.** Family upholds wealth preservation and skepticism against SRI.


#### *4.4. Proving Profit to Conform with Family Norms*

Our data shows that HNWIs are engaged in proving the economic profitability of SRI to conform with family norms, suggesting that a "good" investor is an economically successful investor. This, however, differs from the above-described striving for financial return in that HNWIs primarily aim for economic profit to prove their conformity with family norms. We detail the two aspects related to this insight below.

*Profit to legitimize SRI to the family***.** HNWIs often mention financial success as a source of legitimacy. Firstly, the informants said that *financial gains prove to the family a serious investment strategy*. For example, an interviewed HNWI (HNWI 15) explicates how generating financial returns built the necessary approval from the family hedge fund for adopting an SRI strategy: "my hedge fund, this email I got was, 'oh it (SRI) sounds just like charities, and no problem, you can be on the board'". However, this wealth owner seeks to demonstrate that SRI is not charity, but allows for the generation of financial gains, to convince family members that SRI is a serious investment strategy: "they (members of the family hedge fund) approached me to be on the board, but it's actually not okay like I

want people to realize that it can be very profitable, and it is important for me to generate returns so that again you can prove this concept".



Secondly, our interview partners render financial return and the proof of *profitability as the vital reference point for family members* and a known and appreciated measure for assessing individuals within the family. Suppose the individual HNWI can provide evidence that an investment decision generates enough profit. In that case, influential family members, such as the grandfather, acknowledge this as sufficient to let the individual (i.e., in our example here, the grandchild) proceed with their own ideas. It thus justifies the position of a capable, independent decision-maker. This is illustrated in the following statement by a wealth owner (HNWI): "I decided to talk to my grandfather, and I told him that I wanted to work with education and that it was something that would change the world. The only thing that he said was, 'but how are you going to pay your bills?'".

In a nutshell, the interviewed HNWIs indicate that they use financial success as a source for legitimizing SRI to their family members. This approach is explained on the one hand by the circumstance that HNWIs draw on economic profits to prove a serious investment strategy; for example, to receive approval from their family hedge fund for adopting an SRI strategy. On the other hand, financial success is the vital reference point for assessing family members and thus for whether an individual family is considered appropriately competent to invest the family capital in SRI. In Table 8, we provide further evidence of financial success as a means of legitimization within the family.

*Making profits to achieve recognition*. SRI-oriented HNWIs strive to gain recognition as investors, for example, from their families, by making financial profits. Firstly, nextgeneration wealth owners born into their societal position point out that they need to find ways to show that their actions are credible. A ubiquitous way to achieve this goal is profit because *financial gains increase credibility*. One interviewed HNWI (HNWI 5) describes the

importance of bringing proof to the family as a financially successful investor using the following comparison: "you're expected to shape your life so that you can become a good steward (of your inherited wealth), versus, 'oh I have this, great, I just found out, so I don't have to work as hard, I don't have to find a job, I can just rely on my family'". Thus, recognition in the family is obtained by distinguishing oneself as a financially successful steward of inherited wealth.

Secondly, because wealthy sustainable investors often consider making profits essential for achieving recognition, they usually suggest that the *social benefit is secondary to profit*. HNWIs often do not show their ambition to prove the impact case of SRI to meet the initial intention of a social or environmental purpose. One interviewed HNWI (HNWI 31) illustrates this by pointing out that the measurement of any positive social or environmental impact merely distracts from the central goal of making a financial profit: "this whole discussion about impact measurement, I think, is diverting maybe too much resources from thinking about how to make this financial success first". Hence, in the case of SRI engagement, the social benefit is systematically subordinated to financial profit.

**Table 8.** Profit to legitimize SRI to the family.


In summary, our analysis of the interview data suggests that SRI-oriented HNWIs strive to gain recognition as investors from their family members by making financial profits. This insight is evidenced first by HNWIs aligning their investments primarily with financial performance to make their actions more credible, and second by making the measurement of any positive societal impact secondary to proving financial performance. In Table 9, we provide further evidence for the role of making profits to achieve recognition.

#### *4.5. Other SRI-Oriented HNWIs Provide Confirmation*

The HNWIs in our data frequently pointed out other SRI-oriented HNWIs whom they admire and who serve as a reference for them. We detail two aspects related to this insight in the following.

*Sharing one's own goals and vision with other HNWIs*. Our informants often praise the community of other SRI-oriented HNWIs and how they thrive on being surrounded by like-minded private investors who share their goals and visions. Firstly, other SRI-oriented HNWIs are necessary for a wealthy sustainable investor to *exchange ideas about tackling specific issues from an SRI perspective*. An investment advisor (Expert 12) who regularly consults with HNWIs further elaborates on this very issue by pointing out the relevance of "a community of like-minded investors". Such a community allows SRI-oriented HNWIs "to deep-dive into a specific issue area" and how to "tackle that from a sustainable investing standpoint".

**Table 9.** Making profits to achieve recognition.


Secondly, our informants frequently emphasize the importance of *learning from other HNWIs*. One HNWI (HNWI 25) explains the importance of learning from others in the context of a global network of impact investors as "being part of a more global community of impact investors was extremely helpful." According to the informant, this worldwide network of SRI-oriented HNWIs derives its importance, particularly in representing a community, "from that you can learn".

To sum up, the interviewed HNWIs point out the relevance of sharing their goals and visions with other SRI-oriented wealthy private investors. This relevance stems from the fact that like-minded investors provide an individual HNWI with the opportunity to share ideas on approaching specific issues from an SRI perspective and learn more about SRI from other HNWIs. In Table 10, we provide further evidence for the role of sharing one's own goals and vision with other HNWIs.

*A community of values with other SRI-oriented HNWIs*. In contrast to their families, other HNWIs do not demand anything from our informants. While family members claim their demands for a financial profit, other SRI-oriented HNWIs do not make any demands, either in terms of economic gain or contribution to social welfare. Firstly, this becomes evident by the circumstance that *actual SRI investment skills are irrelevant* to participation.

One wealth owner (HNWI 26) accordingly points out that every HNWI is welcome to the community of SRI-oriented HNWIs regardless of where the person is on the SRI journey: "it's very nice to be welcomed by a group that says, 'if you want us to support you on your journey,' that term is used a lot, the impact journey that we're on here". Thereby, it is more about experiencing the journey toward making a positive social impact with a group of like-minded SRI-oriented HNWIs than actually about achieving the goal of creating a positive impact. "I don't feel as pressed to come up with something perfect, but rather to have a full journey with a group of like-minded individuals" (HNWI 26).

**Table 10.** Sharing goals and vision with other HNWIs.


Secondly, our informants frequently mentioned that *sharing similar values connects one to another*. One informant (HNWI 10) clarifies the importance of being surrounded by like-minded HNWIs who share the same goals and visions and how such a community serves as a source for inspiration and support because "you will feel alone, and also, you will not be able to scale if you are alone ( . . . ). And here comes a certain belief, that of conviction." The shared set of values among SRI-oriented HNWIs creates a sense of community, which is a crucial source of guidance for the individual wealth owner. In fact,

according to the same informant, "it's always important to be a part of a community that you share with a grandiose ambition".

In conclusion, our analysis of the interview data indicates that other SRI-oriented HNWIs serve as a community of values that does not impose concrete requirements on an individual HNWI, neither in terms of financial gain nor of positive social impact. Namely, on the one hand, whether an individual HNWI has SRI skills and thus actual knowledge of how to link economic and social aspects is irrelevant to belonging to the community of SRI-oriented HNWIs. On the other hand, as a community of values that does not impose any concrete requirements on an individual HNWI, it is mainly about sharing the same goals and visions. In Table 11, we provide further evidence of a community of values with other SRI-oriented HNWIs.

**Table 11.** A community of values with other SRI-oriented HNWIs.


#### **5. Discussion**

#### *5.1. The Influence of Reference Groups on the Investment Behavior of SRI-Oriented HNWIs*

While we know little of the investment behaviors of SRI-oriented HNWIs, reference group theory suggests that such behavior is centrally dependent on their identification with and comparison to a respective reference group. For this reason, we have set the objective of developing knowledge on the influence of reference groups on the SRI engagement of HNWIs. Based on an inductive qualitative investigation of 55 semi-structured interviews with HNWIs and industry experts, we developed a framework to explain how reference groups influence the investment behaviors of SRI-oriented HNWIs. Our framework indicates that the family directly influences and other SRI-oriented HNWIs indirectly influence

SRI-oriented HNWIs towards generating financial profits in their investments at the expense of social welfare considerations. On the one hand, the family serves as a normative reference group that upholds the economic profit motive and directly urges HNWIs to make financial gains from their investments at the expense of social welfare. On the other hand, other SRI-oriented HNWIs serve as a comparative reference group that shares the same values but does not impose any concrete requirements on social welfare performance. This indirectly influences SRI-oriented HNWIs to subordinate social concerns to financial profits. Figure 2 provides an overview of our explanations.

**Figure 2.** How reference groups influence the investment behavior of SRI-oriented HNWIs.

Our framework shows that SRI-oriented HNWIs are open to the idea of combining social welfare and economic aspects in their investments (see the two boxes with dashed and solid lines at the bottom of Figure 2). On the one hand, they intend to use their own fortune to promote social welfare. SRI-oriented HNWIs regard wealth both as a cause for the imbalance between rich and poor and a solution to overcome this very inequality. The latter explains the personal responsibility HNWIs ascribe to contributing to social welfare by placing their wealth into SRI. On the other hand, HNWIs intend to use their own fortune to generate financial profits. They regard financial return as essential, considering SRI as a financial vehicle to contribute to social welfare but also to make an economic profit. Moreover, HNWIs argue that financial gain serves the cause of SRI, considering profitability as a prerequisite for spreading SRI amongst mainstream investors.

However, while SRI-oriented HNWIs are open to the idea of combining social welfare and economic aspects in their investments, they strive towards making a financial profit at the expense of social welfare considerations even though they already hold great fortune (see the box with the solid line at the bottom of Figure 2). The influence of two particular reference groups explains this profit-oriented investment behavior of wealthy private investors.

First, a push-and-pull effect between the family setting profit-oriented norms and the HNWIs proving profit to conform with family norms directly promotes SRI-oriented HNWIs' ventures for financial return (see the box at the top right and the corresponding vertical arrow in Figure 2). The push consists of the family that serves as a normative reference group [23], setting profit-oriented norms that wealthy sustainable investors perceive as standards and expectations they must adhere to. Family members tend to have traditional investor mindsets, suggesting that lent or invested capital needs to generate financial profits to compensate the risk that the investor takes by giving the money away. From this normative group's perspective, the only reasonable explanation for taking such a risk is a financial profit. Consequently, the family upholds the value of wealth preservation and skepticism against SRI and suggests that SRI does not fall within the purpose of the family. The pull is that SRI-oriented HNWIs strive for financial profit to conform with the norms of their families, upholding the importance of economic profits. They try to make profitable investments to legitimize SRI to their family members and to achieve their recognition. However, by these activities, SRI-oriented HNWIs reinforce and further consolidate family norms, countering the underlying idea of SRI, which brings together financial profits and social welfare (e.g., [6]).

Second, other SRI-oriented HNWIs provide confirmation and thereby indirectly promote SRI-oriented HNWIs' ventures for financial profits (see the box in the middle and the corresponding horizontal arrow in Figure 2). These like-minded individuals allow SRI-oriented HNWIs to share goals and vision with their peers and serve as a community of shared values. Within this group, an SRI-oriented HNWI finds validation for own ideas of using financial capital for social welfare and acceptance that the consideration of ESG criteria is appropriate and reasonable. In this vein, other SRI-oriented HNWIs build a comparative reference group, as they serve as a standard or checkpoint which the individual uses to make judgments [23]. However, this reference group does not enforce any standards, as can be seen, for example, in that actual SRI investment skills are irrelevant for membership. Consequently, those judgments are decoupled from the investment behavior of SRI-oriented HNWIs. For this reason, the comparative reference group has at least an indirect positive effect on profit-oriented investing by reinforcing the influence of the normative reference group on the profit-seeking of SRI-oriented HNWIs.

#### *5.2. Contributions to the Literature*

Our study adds to SRI research. To achieve sustainable development, we need a shift of traditionally invested assets into SRI. HNWIs hold a vital role in this shift, controlling nearly half of global wealth [3]. However, we know little about wealthy sustainable investors [16,17] and SRI-oriented HNWIs [18,19]. To understand whether, how, and to what extent HNWIs engage in sustainable investing, we need to go well beyond whether or not SRI is more profitable than traditional financing because the former brings together financial profits and social welfare [6,9]. We showed that SRI-oriented HNWIs use their fortune to generate economic gains at the expense of social welfare in their investments and unpacked the reasons behind their profit-oriented investment. While they support the idea of mobilizing their wealth to promote social welfare, they let this goal fall short because of reference groups that encourage them to use their wealth to generate financial profits, even though they already hold great fortune. The insight that the SRI engagement of HNWIs is, in effect, primarily profit-driven due to the direct influence of family members and the indirect effect of other SRI-oriented HNWIs, suggests that such engagement could contribute less to social welfare and more to further boosting wealth inequality. This finding is accentuated by the COVID-19 pandemic, which has again exacerbated existing wealth inequalities [61].

We further contribute to the reference group theory literature. As mentioned above, the literature differentiates two types of reference groups [22–24]. While normative reference groups establish and enforce standards considered norms, comparative reference groups serve individuals as a point of reference in making evaluations or comparisons without the evaluation of the individual by others in the group. By focusing on how different reference groups influence the investment behaviors of SRI-oriented HNWIs, we can comparatively show how different reference groups each affect the profit and welfare orientation of wealthy investors. This lets us derive an exciting finding for reference group theory. In the case of conflict, normative reference groups suppress the beliefs, values, and perspectives of the comparative reference groups. Suppose that the reference group does not enforce its values or does not even seek to do so. In that case, this space is occupied by a reference group that does, while the comparative reference group at least indirectly supports the standards of the normative reference group. This insight implies the different spheres of influence of normative and comparative reference groups. In addition, understanding how different reference groups influence values, which then, in turn, shape the investment behaviors of SRI-oriented HNWIs, echoes the relevance of values for studying contexts where, as in the case of SRI, it is a matter of conceptualizing the interactions between economic issues and social aspects [46,62].

#### *5.3. Contributions to Practice*

The knowledge gained into the influence of reference groups on the investment behaviors of SRI-oriented HNWIs demonstrates that it is critical for market participants to be highly aware of the specific social setting that their HNWI clients or constituents are in when they receive their messaging. That is because the social setting in that moment will serve as a critical contextual aspect in determining what types of arguments about SRI—financial or social welfare arguments—will resonate more or be more helpful for HNWIs to move ahead with an investment decision. More specifically, in a shared ownership setting, as in families, financial arguments are more likely to support an investment decision. In contrast, social welfare arguments are more likely to support an investment decision in the setting of an SRI-interested HNWI community.

For the managers and members of communities of SRI-interested HNWIs, our findings suggest that to drive the primary goal of social welfare more effectively and to overcome the dominance of the financial performance-seeking of other family members, it might be crucial to put more specific emphasis within their community on the actual achievement of social goals, to drive more specific goal-setting in that regard, or to set certain standards and minimum requirements within their community.

Our research insights point out that mobilizing private wealth, at scale, for a positive social impact requires a deep understanding of the underlying social contexts that HNWIs are embedded in and which substantially influence their investment decision-making. Specifically, for the crucial intermediaries of banks and SRI funds, our findings indicate that to mobilize private wealth into SRI products, it is relevant for financial intermediaries to carefully consider and shape the social settings in which their HNWI communication activities occur. Depending on the settings of their specific activity, either financial or social welfare arguments might impede, rather than support, unlocking the substantial latent demand for their SRI products. It is these social setting considerations, and them not being considered carefully, that so far might have been the crucial stumbling block for SRI in private wealth management.

#### *5.4. Limitations and Future Research*

Our research is not without limitations—many of which are linked to its qualitative nature (see [63]). However, we believe that it opens up a broad range of future research opportunities that can add nuance and clarity to the possibilities and limitations of HNWIs' contribution to a shift of traditionally invested assets into SRI and the influence of different reference groups in this process. Our qualitative research strategy allowed for more accurate insights into the context of SRI-oriented HNWIs' investment behaviors, which would have been challenging to obtain through quantitative approaches. However, this also means that qualitative research develops generalizations that "are often less parsimonious because of the large number of variations possible and the difficulty of predicting which ones will occur and why" [64] (p. 703). Future research could use a quantitative method to test the generalization of our study statistically and enrich the boundary conditions of our work—for instance, linked to geographic or personal aspects.

While our data allowed us to theorize the influences of different reference groups on the investment behavior of SRI-oriented HNWIs, more research is needed to examine the gradual transition of these influences and potential shifts in them over time. Longitudinal

studies could further decipher the temporal dynamics behind the influences of reference groups on individual investment behaviors and any measures individual investors take to counter the influence of third parties. Examining the influence of such groups over different points in time could explain how and why a particular group manages to assert itself over others, what the associated influence strategies are, and why they are particularly assertive with the respective investors. Such research could also reveal whether, how, and why investors evade the influence of third parties and what the respective preconditions are for escaping the influence of a particular reference group (e.g., social embeddedness, individual strategies against influence). In addition, future research could examine the individual capabilities of private investors who positively impact social welfare through their investments, even in a context where financial gain is preferred over social welfare engagement.

#### **6. Conclusions**

A reference group theory perspective suggests that SRI-oriented HNWIs' investment behavior is shaped by their identification with and comparison to reference groups. To close the existing knowledge gap regarding HNWIs' SRI engagement, we adopted a qualitative interview approach to examine how reference groups influence the investment behaviors of SRI-oriented HNWIs. We found that the family members of SRI-oriented HNWIs form a normative reference group that prioritizes financial returns and directly shapes HNWIs to subordinate social concerns to financial profits. Our study also indicated that fellow SRI-oriented HNWIs serve as a comparative reference group that does not impose any concrete requirements on social welfare performance, indirectly influencing SRI-oriented HNWIs to generate financial gains from their investments at the expense of social issues. Our scholarly insights contribute to the SRI literature and reference group theory and have practical implications.

**Author Contributions:** Conceptualization, D.R., F.P. and A.K.; methodology, D.R., F.P. and A.K.; software, D.R. and A.K.; validation, D.R., F.P. and A.K.; formal analysis, D.R. and A.K.; investigation, D.R., F.P. and A.K.; resources, D.R. and F.P.; data curation, D.R., F.P. and A.K.; writing—original draft preparation, D.R., F.P. and A.K.; writing—review and editing, D.R., F.P. and A.K.; visualization, D.R. and F.P.; supervision, not applicable; project administration, D.R.; funding acquisition, not applicable. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


*Article*
