Corps de l’article

Over the last two decades, microfinance has been receiving increasing global acceptance as a social inclusion mechanism and a poverty-alleviating development strategy (Easterly, 2006). What started as small-scale, isolated, microlending experiments rapidly became a fast growing, globalized practice. Among several definitions of microcredit found in the literature, we adopted that which refers to “programs that extend small loans to poor people for self-employment projects that generate income” (Woller and Woodworth, 2001). As microfinance institutions (MFIs) began to provide a greater variety of financial services (e.g., micro-insurance, micro-savings), the term microfinance was coined. In the definition used throughout this paper, microfinance is related to the access and use of financial services by a low-income population, mainly those typically excluded from the traditional financial system. Therefore, microcredit is one component of microfinance, probably the most prominent but not necessarily the most important (Dichter and Harper, 2007).

In general, MFIs provide loans through two basic methodologies: individual and group lending. The individual lending approach consists of providing loans to individuals, usually requiring some sort of collateral (and sometimes a co-signer) as a guaranty for the loan. The group lending approach uses “social collateral”, i.e., it allows the poor to obtain loans without collateral, their peers serving as guarantors (Mitra, 2007). The members are selected by their peers, who also evaluate their trustworthiness (Khan and Rahaman, 2007). Regularly scheduled meetings allow members to collect repayments and savings. A common point between group lending and individual lending is the fact that they both rely on the client’s social network to verify his/her trustworthiness before disbursing a loan (Jayo et al., 2009). The group-based lending approach comprises a number of specific methodologies, including two that are particularly relevant to this study: solidarity groups and village banks. The main differences between these two approaches are group size (solidarity groups number 3-10 people while village banks usually number 20-50), the market segment (unlike solidarity groups, communal banks target the poorest clients) and loan size (targeting poorer clients, village banks lend smaller amounts).

Following the success of microcredit organizations such as the Grameen Bank in Bangladesh, Bank Rakyat in Indonesia, and Banco Solidario in Bolivia (Woller and Woodworth, 2001; Dorado-Banacloche, 2001), wealthy nations and international agencies have provided billions of dollars to support microfinance projects. They eventually started to finance projects of replication of these successful models and their respective methodologies and tools in order to expand microcredit “best practices” around the world. However, these “replicating” initiatives did not always provide good results. In fact, scholars and practitioners realized that the host countries often needed to significantly adapt the original methodology for it to be implemented sustainably in their own context (Hulme, 1990). In other cases, the replication and implementation of the foreign model simply could not be achieved (Seibel and Torres, 1999).

A number of scholars undertook to investigate the question of microcredit methodology replication, suggesting that the emergence of new MFIs – and consequently the development of the microfinance sector – might depend on these institutions’ capacity to appropriate successful models for their own context (Nichter et al., 2002). The general question guiding this empirical study is: What are the conditions facilitating replication of microcredit methodologies? In order to explore this more general question, we decided to investigate the conditions facilitating the replication of one of the most successful Brazilian experiences, the CrediAmigo methodology. CrediAmigo is a microcredit program implemented successfully in the Northeast of Brazil and then transferred to Rio de Janeiro. We believe that by investigating this particular case of microcredit replication we can shed light on conditions that could also be valuable in other contexts.

Microfinance in Latin America and Brazil

There exist four types of MFIs in Latin America: NGOs, for-profit MFIs, commercial banks and finance companies. In the 1980s, NGOs started to offer subsidized microcredit. In the early 1990s, a number of NGOs became for-profit MFIs and others became commercial banks. From the end the 1990s, new banks began to enter the market, increasing competition in the sector (Marulanda and Otero, 2005). In addition, finance companies, which lend credit for consumption, have also become strong market competitors (Christen and Cook, 2001).

The current situation of microfinance in Latin America is not homogeneous. According to The Economist (2009), that evaluated microfinance as a commercially viable and sustainable activity using thirteen distinct indicators, Peru and Bolivia (along with the Philippines) are at the top, leaving behind more than fifty five countries. Other Latin American countries among the well ranked are Ecuador, Nicaragua, Colombia and El Salvador. Larger countries, like Brazil, Mexico and Argentina, are not well ranked in the microfinance market. However, it is important to point out that these rankings are an attempt to evaluate countries with databases that can be often considered poor and incomplete.

For a complete view of the evolution of Brazilian microfinance market since the 1970’s, see Monzoni (2006). As microfinance gained popularity worldwide during the 80’s and 90’s, Banco do Nordeste do Brasil (BND), a regional development bank, came to perceive that microfinance was a significant opportunity for economic development of the Northeast of the country, one of the poorest regions in Brazil (Souza, 2008). Hence, BND approached the World Bank in 1996 seeking support for implementation of a microcredit program. At the time, the World Bank was pushing for development of microcredit initiatives in various countries (Brusky, 2003).

BND needed to gain knowledge of best practices in microcredit, so managers were sent to visit five different MFIs in Brazil, Bolivia, Chile, Peru and Indonesia. Following these visits, BND sent managers to Banco Solidario (Bolivia), a success-story in Latin-American microfinance that was originally inspired by the renowned Grameen Bank[2]. In November 1997, the CrediAmigo program was created. BND set up five pilot agencies in capital cities and smaller towns, proposing a standardized financial product, called CrediAmigo Solidario, involving groups of five to seven members, three-month loans and six fortnightly payments. One of these pilot projects was located in Fortaleza (capital of Ceará, a state in Northeastern Brazil). The program’s rapid expansion in 1998-99 resulted in payment default problems, which impacted the portfolio quality. In response, BND decided to develop a solid microcredit monitoring process, managed by a well trained team (Brusky, 2003).

From 2004 to 2010, the program grew significantly in terms of number of clients and active portfolio (total amount lent). CrediAmigo currently has more than 170 units distributed essentially within the Northeast and reaching more than 650,000 clients. Since the beginning of 2009, CrediAmigo has expanded its operations to Rio de Janeiro, located in the Southeast of Brazil, a quite different region in terms of socio-economic conditions and development level. This explains why we focused our investigation on the process of transferring CrediAmigo from the Northeast, particularly in the state of Ceará, to the Southeast, particularly the state of Rio de Janeiro, a richer and more developed Brazilian region.

Microcredit methodologies’ replication

Since the mid 1980s, numerous replications of successful microcredit models have been carried out to widen the outreach of microcredit worldwide. Still, there are few studies on this issue. Our review of literature discussing microcredit methodologies replication was organized according to three basic questions: what is replicated, how it is replicated and where it is replicated.

What is replicated? Microcredit methodologies replication implies certain core principles (basic features of the methodology) that should be transferred from one context to another. Gibbons (1999) and Siebel and Torres (1999), analyzing the Grameen Bank, proposed three core principles: (1) a high leadership commitment focusing exclusively on the poor – particularly poor women; (2) the provision of financial services facilitating participation and repayment; (3) the quickest attainment of financial self-sufficiency through credit discipline. However, these fundamental features cannot be simply copied from one institution to another due to different environments, and thus need to be adapted.

How is it replicated? Hossain (1988) and Hulme (1990) have analytically categorized four possible scenarios for replication of the Grameen model. First, the model’s organizational framework is considered replicable on a national scale in its current form. Second, the original model can be used as a starting point for a small-scale action research project, which can eventually be modified to suit a new context. Third, an organization may adopt a number of features and policies pertaining to the original model and reject others. Fourth, the model is appreciated as an interesting case but little can be transferred to another institution. These scenarios suggest that, the more the two contexts differ, the more stakeholders need to adapt the original model.

Where is it replicated? The conditions required to successfully replicate a microcredit methodology are institutional (what is under the control of the institution) and societal (the institution’s environment) in nature. Examples of institutional conditions are access to sufficient funds to support the institution’s initial expansion and consolidation (Seibel, 1998) and a well-motivated, committed and trained staff (Dick, 2002). Among societal conditions we found economic conditions (a healthy economy opened to free market) (Boily, 1999), political conditions (political stability and freedoms, and stakeholder support) (Gibbons, 1999), cultural conditions (recognition of usury and credit, and of women’s right to credit) (Renteria, 1996) and legal conditions (clear and transparent banking laws).

Although the replication of microcredit methodologies has received increasing attention, much more work is needed in the area (Diniz et al., 2009). To our knowledge, no author has studied the replication of microcredit methodologies in the Brazilian context. This served as a primary motivation for our study.

Conceptual framework

The conceptual framework adopted in this study is based on a recent constructivist framework proposed by Pozzebon et al. (2008), which has been conceived to guide empirical investigations at the community or societal levels of analysis and that has been increasingly adopted by microfinance researchers in Brazil (e.g., Jayo, 2010; Cernev, 2010; Diniz et al. 2009). We believe that such a framework, which postulates that any social research is processual and inherently multilevel, was appropriately adapted to guide our study. The framework is based on two theoretical perspectives: the social shaping of technology (SST) and the structurationist view of technology.

Work on the social shaping of technology (SST) has been greatly influenced by the sociology of knowledge (Berger and Luckmann, 1967) and the social construction of technology (Bijker and Law, 1992). Aiming at overcoming the deterministic concepts of technology frequently found in the management literature, the SST perspective allows considering the implementation of a particular technology as a negotiation between social actors. Such a view aims at identifying occasions in which decisions and actions related to management of technology and change can effectively be applied (Clausen and Koch, 1999). Three SST concepts were selected to integrate the framework: (1) relevant social groups, (2) interpretive frames and (3) the technology implementation as a negotiation process.

The first concept – relevant social groups – refers to a group of individuals pertaining to a common functional, geographical or virtual space who share a number of values and positions on a particular subject of interest. Subgroups and alliances between these groups may lead to the creation of social spaces and play an important role. In this paper, the relevant social groups considered are those related to the use and application of the microcredit methodology under investigation, including managers and coordinators of the microcredit program, credit agents, and clients.

Interpretive frames, the second concept, are based on the premise that individuals taking part in a relevant social group are likely to share a set of assumptions, values, positions, beliefs or objectives related to a given subject. In our work, it refers to the application, replication and adaptation of a particular microcredit methodology. The group’s members are likely to share similar perceptions and expectations with regard to a particular microcredit methodology.

The third concept – negotiation process – refers to the process of transferring, adapting, or simply replicating a given microcredit methodology. Different groups will bring differing interests, stakes, and visions to that negotiation process, which will influence the process and the results of the negotiation (Clausen and Koch, 1999). In this research, we stress the negotiation process between the relevant social groups involved in the replication of a particular microcredit methodology.

The second theoretical perspective constituting the multilevel framework is the structurationist view of technology, inspired by structuration theory[3], which represents an important stream of thought in information systems and management research (Pozzebon and Pinsonneault, 2005). Of the most influential researchers employing this theory is Orlikowski (2000), who developed a seminal concept that addresses the role of emergence and improvisation in technology-related implementations, called technology-in-use or technology-in-practice. We adapted this concept to formulate the concept of methodology-in-practice.

The concept of methodology-in-practice will focus on the consequences of the implementation and use of a particular microcredit methodology. In this research, methodology-in-practice refers to the consequences or results of the process of redefining and adapting a microcredit methodology to a new context.

Figure 1 illustrates our framework, beginning with indication of the three levels of analysis: individual, group and community. The group level represents the concept of relevant socialgroups, which bridges the gap between micro and macro levels: individuals and their communities, neighbourhoods or cities. To each relevant social group, we have attached the interpretive frames, i.e., their assumptions and expectations regarding the implementation and use of a given microcredit methodology. The negotiation process consists of identifying the occasions and spaces by which relevant social groups, interacting over time, replicate a given methodology. The results of the negotiation process are represented by the methodology-in-practice, i.e., the way the methodology has been adapted and its consequences for the social groups involved.

Figure 1

The multi-level framework guiding the research

The multi-level framework guiding the research

-> Voir la liste des figures

Our framework is considered “multi-level” because it combines inextricably linked levels of analysis (individual, group, and local community), blurring the line between macro- and micro-level interactions, considering that changes at the community or societal level are initiated at the individual or group level and vice-versa.

Methodological approach

This research aimed at understanding the conditions for local appropriation and transferability of microcredit methodologies. In order to explore the research question, we applied a case study method (Stake, 1998). The selected case was the CrediAmigo program, hosted by Banco do Nordeste (BND). CrediAmigo is considered one of the most promising Brazilian microcredit institutions to date. First, we studied the organization’s microcredit delivery model, which was implemented in 1997-8. Second, we investigated the conditions and challenges as perceived by the social actors involved in the process of transferring and adapting its methodology to other regions of Brazil, particularly from Ceará to Rio de Janeiro.

For the purpose of this investigation, we relied on an instrumental single-case design (Stake, 1998). According to Stake, a case study is instrumental when the case will be examined in order to advance knowledge about a particular topic under scrutiny. The rationale supporting the choice of BND’s CrediAmigo is that this program is the only microenterprise microcredit program to have reached significant scale in Brazil and to have already been successfully transferred from a foreign context to the Brazilian one (CrediAmigo was inspired by the Bolivian Banco Solidario, which in turn was inspired by the Bangladeshi Grameen Bank).

Data collection: The information collected during the research came from three complementary sources: interviews, documentation and observation. During six months, the first author was in immersion on the site, carrying out real-time observations, interviews and collecting documents. We identified four main relevant social groups: program managers, coordinators, credit agents and clients of CrediAmigo. CrediAmigo is structured in three levels: regional agencies, local agencies and local units. The program managers supervise the management of regional and local agencies. The coordinators are responsible for specific agencies or units and are in charge of all of the credit agents involved with each of the local agencies or units. The credit agents are in direct contact with clients and they are responsible for a number of procedures such as clients’ first contacts, credit analysis, coaching, etc. Finally, the clients are those low-income persons benefiting from microcredit loans. Fifteen respondents were selected according to the concept of relevant social groups (Table 1). All the names of the interviewees were changed to guarantee anonymity. The duration of the interviews varied from 30 to 1:30 minutes in average.

Table 1

Presentation of the interviewees

Presentation of the interviewees

-> Voir la liste des tableaux

The criteria for selecting interviewees were basically three: (1) their belonging to one of the relevant social groups; (2) the richness of information they were willing to provide, based on a snowball process of suggestions and recommendations by BND managers; and (3) their availability and willingness to participate in the interviewing process. Our initial target was at least three respondents from each relevant social group. The only group for which we have not attained our goal is the coordinators. However, we can attest that the two coordinators interviewed demonstrated a solid and rich view of the conditions for the replication of CrediAmigo, the result of extensive experience in the institution. We interviewed four managers working with business strategy, i.e., involved with the replication of the program in other regions, two coordinators in charge of the management of specific agencies or units, three credit agents who work on the field with the communities, and six clients of the program. The data obtained were integrally transcribed. It is important to note that a research protocol (in Portuguese) was built based on the research framework. The protocol encompasses interview guidelines as well as ethical-related forms, like letters of consent.

Data analysis: We proceeded with the analysis of empirical material following the method proposed by Miles and Huberman (1994). This approach to data analysis comprises three phases: data condensation, data presentation and elaboration and verification of conclusions. The four concepts of the framework were used for “analytical induction”, i.e., they were used to guide the initial categorization while leaving however space for the identification of new categories and the refinement of existing ones to enrich our understanding of the studied phenomenon (Patton, 2002).

Analysis and discussion of the results

In this section, we present the results of our data analysis in terms of the main concepts of our theoretical framework: interpretive frames, negotiation process and methodologies-in-practice. In order to build a plausible answer to the research question, we compiled and compared the perspectives that emerged from the four relevant social groups – managers, coordinators, credit agents and clients – concerning the conditions facilitating replication of the CrediAmigo methodology.

Interpretive frames: the most important principles of the methodology

In order to capture the respondents’ interpretive frames, we asked the following question: what does the interviewee consider the most important principles of the methodology. Table 2 shows the most significant principles that emerged from the interviews, in order of importance and by relevant social group.

Table 2

Most important principles of the methodology, by relevant social group[4]

Most important principles of the methodology, by relevant social group4

-> Voir la liste des tableaux

Two principles emerged as more important among the essential principles of CrediAmigo methodology, which are relevant for a better understanding of the conditions facilitating its potential replication. They are: (1) commitment of the credit agent and (2) peer selection and enforcement.

Commitment of the credit agent to his/her job was the first aspect considered by 9 respondents. A committed credit agent would be one who feels s/he has a mission within the organization (alleviating poverty, assisting people) and who believes in his/her work tool, the microcredit methodology.

The most important is to have committed individuals, who understand what they are doing, who perceive something beyond the provision of credit (...) who have been provided with adequate training (...) I need to have the methodology, but I [also] need individuals who believe in that methodology, who have internalized that methodology.

Sténio, coordinator

Peer selection and enforcement refers to mechanisms that consist of transferring the responsibility for any client’s payment default to his/her peers within the group. Here, peers refer to members of the lending group. Peer selection and enforcement can be considered a social filter in that people choose those among them who are most trustworthy and thus should be allowed in the group. This implies a minimum of solidarity and respect between peers.

Peer selection is appreciated because it determines whether the group will be good or not. If I set a group of four individuals, it is three people telling the bank that I can be trusted. The larger the group, the higher will be the level of trust that the bank will maintain in relation to that group. This is because if you have a group of ten people, it is nine of them saying they can trust the tenth. (…) I think that this is the most interesting part of the methodology.

José, credit agent

Comparing these results with the literature, we recall the analysis carried out by Gibbons (1999) and Siebel and Torres (1999). They have suggested three essential principles of the Grameen methodology: (1) high commitment of the leadership focusing exclusively on the poor – particularly poor women; (2) provision of financial services facilitating participation and repayment; (3) quickest attainment of financial self-sufficiency through credit discipline. The comparison produced three main conclusions.

First, the Grameen’s high commitment of the leadership focusing exclusively on the poor does not appear to be a priority in the case of the CrediAmigo program. This may be due, at least partly, to historical reasons. CrediAmigo was not set up by a leader dreaming of getting his people, and particularly women, out of extreme poverty by setting up a financial organization devoted to serving the poorest of the poor. Instead, CrediAmigo was set up by an experienced public development bank (BND) which perceived microfinance as a potentially sustainable socio-economic development strategy. What emerged from CrediAmigo’s analysis was the importance of the commitment of the creditagent, which we suggest as a contribution to existing literature.

Interestingly, our results corroborate the findings produced by Gibbons (1999) and Siebel and Torres (1999) regarding the second Grameen principle – the provision of financial services facilitating participation and repayment. When compared to peer selection and enforcement, we concluded that both are related to creating conditions for achievement of the same goal: the long term sustainability of the lending group and of the MFI (in this case, CrediAmigo and Grameen) itself.

Third, we may suppose that the quickest attainment of financial self-sufficiency was not as much of a priority for CrediAmigo as it was for the Grameen Bank, since CrediAmigo was implemented within an experienced sustainable financial institution which already had a significant portfolio in other areas when it was created. However, the bank decided that its first product (CrediAmigo Solidario) would not focus exclusively on the poor, in order to facilitate the attainment of financial self-sufficiency. This shows that financial self-sustainability was indeed a priority for the bank, even though not the most urgent one. In addition, the product that would allow the program to reach even poorer clients (CrediAmigo Comunidade) was implemented about seven years after the initial implementation of the program. This means that the program had time to reach self-sustainability before aggregating products that provide less return. Hence, the institution could focus energies elsewhere, namely training committed credit agents and developing a monitoring process aimed at ensuring that credit agents do their job well and detecting any problem at an early stage.

Finally, we performed a critical analysis of the different views of coordinators (CO), managers (MG) and credit agents (CR). While the groups managing the MFI – managers and coordinators – provided a more global view of the methodology and program, the credit agents provided an understanding that is closer to ground level, i.e., the clients. This certainly does not mean that the perception of one group is worth less, since it can bring in insights not considered by others. In this case, we found a rich complementarity among the different perceptions.

Negotiation process: the challenges for CrediAmigo’s replication

In order to gain a better understanding of the negotiation process that takes place during the implementation and use of CrediAmigo methodology, we questioned managers and coordinators on the following question: what does the respondent expect to be the main challenges related to the replication of CrediAmigo?

Table 3 shows the main challenges for replication according to different relevant social group perceptions (we retain in the Table just the most relevant factors, i.e., those that emerged with greater frequency and plausibility from the voices of the respondents). Although all challenges depicted in Table 3 should merit particular attention in the process of replication of a microcredit methodology, we outline here the first two: (1) establishment of a partnership with a local organization and (2) microentrepreneurs’ resistance to group lending methodologies.

Table 3

Challenges for replication, by relevant social group

Challenges for replication, by relevant social group

-> Voir la liste des tableaux

Establishment of a partnership with a local organization: First of all, the type of partnership that an MFI needs to get into depends on the market where it is to be implemented. In a region where the offer of microcredit services is low, CrediAmigo would not depend on its partner as much as in a highly competitive market. In the former case, the program could deal with a given organisation (NGO, cooperative, civil society association) that had not yet initiated the provision of microcredit services. That organisation could start operating jointly with CrediAmigo, adopting its consolidated model (structure, procedures, methods, policies). In a more competitive market, as in large cities such as Rio de Janeiro and Sao Paulo, CrediAmigo will need to make more adaptations. In those cases, it may be particularly useful that the partner be an experienced organization, which already knows how to deal within these contexts, helping CrediAmigo gain essential knowledge of the local market and provide assistance in making adaptations.

The establishment of a partnership among two mature organizations is never an easy process. Experienced organizations already have their own mission, vision, culture and values. The challenge is to modify/adapt these features in order to make the partnership viable. Considering the success the CrediAmigo model enjoyed in the Northeast of Brazil, it will try to bring the local organization’s model as close as possible to its own model, be it in terms of culture and values, productivity or field experience. Discussing the replication of CrediAmigo in Rio de Janeiro, two program managers explained the challenge related to “imposing” the program’s best practices on the local partner’s model.

So, our challenge is to fix this relation so it works well – each one knowing what the role of the other is, and change the culture. Because when we set up CrediAmigo, we started from zero, so we created a culture. (…) So we will need to adapt, there will be an initial shock, in this respect, that necessity of imposing a higher level of productivity, so the operation pays for itself, so there be a return. I think that this process is our main challenge.

Gilberto, program manager

Microentrepreneurs’ resistance to group lending methodologies: the implementation of CrediAmigo in Rio de Janeiro can be considered a pilot project for its replication in other Brazilian regions. Indeed, the implementation team is confronted with at least some of the challenges CrediAmigo initially faced in Fortaleza. One of these challenges is the microentrepreneurs’ resistance to group-lending principles. When we conducted the interviews, the employees of CrediAmigo expected a certain resistance on part of the population in Rio de Janeiro. Still, in the present case, the resistance may be greater since the motives for resistance are numerous.

First, the people may not accept these principles because they do not know anything about this system and are already used to another methodology, individual lending. The predominance of individual lending in larger cities could actually be considered a challenge for replication per se. As we observed when analyzing the transcripts, even clients of CrediAmigo in the Northeast would like to receive individual loans. It seems that they agree to participate in solidarity groups because they do not have other viable options, at least until they expand their business and demonstrate their trustworthiness. Therefore, it is hard to imagine how microentrepreneurs who can already get access to individual loans would accept to depend on a group (which would sound like taking a step backward). Second, contrary to smaller agglomerations, the citizens of large cities typically adopt a more individualistic attitude, making it harder to convince them to submit to peer-lending groups in order to receive a loan.

Based on our experience, the easiest way to work with solidarity groups, which is the program’s flagship, is in smaller towns. This is because people have known each other for a long time. They were born there. The friendship ties are very strong. In the large metropolitan regions, there always is some difficulty.

Rodrigo, program manager

For these reasons, the resistance of microentrepreneurs in Rio de Janeiro may be greater than in Fortaleza ten years before. Based on these insights, and correlating with the literature review previously presented, particularly the scenarios proposed by Hossain (1988) and Hulme (1990), we propose three perspectives regarding the replication of CrediAmigo in other Brazilian regions.

In the first perspective, it is assumed that replication might not be possible in some particularly challenging areas. For these interviewees, replication may not succeed in these areas, since the program would be confronted with at least some of the challenges it encountered at the moment of its implementation in Fortaleza, such as clients’ resistance to group-lending. In addition, they suggest that the most challenging region for replication will certainly be the Southeast (the most populated region), in view of aspects such as the region’s highly competitive market and the existence of a more individualistic mentality.

From the second perspective it is thought that, in spite of being challenging in some areas, replication will probably be successful everywhere. This second group of interviewees appears more confident regarding the replication of CrediAmigo. Most of the issues they consider as challenges for the replication of CrediAmigo are in fact feasible adaptations.

I don’t see any problem of replication in the process, in the program’s methodology. Because we constantly improve, trying to mould our process to the reality on the field. We are always open to the demand, listening to all the possibilities and necessities that may emerge.

Hugo, program manager

From the third perspective, it is considered that not only could the replication of CrediAmigo be successful throughout the country, but a local adaptation to each context is not even perceived as necessary.

A lot of people (…) report the [popular] support that CrediAmigo enjoys in the Northeast. (…) Since CrediAmigo pertains to BND, they will be proud to participate [in the program].

Leno, coordinator

Comparing the answers of different relevant social groups regarding the negotiation process for the replication, it is clear that managers (MG) and coordinators (CO) share a number of challenges in the process. Managers mentioned all of the coordinators’ elements plus others, probably because they have a wider view of the challenges to which CrediAmigo is confronted regarding its expansion and replication. The coordinators apparently gave more importance to criminality as a challenge in larger cities than did the program managers. However, we could claim that this challenge is more dependent on the context than all the others identified. Managers stressed the importance of market competition and technological hurdles (reaching clients in distant areas) as significant challenges, whereas the coordinators did not mention it. This is probably due, as indicated before, to the type of issues each group is confronted with on a day-to-day basis, which is what they are most concerned with.

Methodology-in-practice: the consequences of the implementation of CrediAmigo

As a way to better understand the consequences of the negotiation process that occurred in the course of the implementation and use of the methodology, we asked the respondents to talk about what they perceive as the consequences of the implementation of the CrediAmigo program. Table 4 shows the consequences of the implementation of CrediAmigo according to relevant social groups.

Table 4

Consequences of the implementation of CrediAmigo, by relevant social group

Consequences of the implementation of CrediAmigo, by relevant social group

-> Voir la liste des tableaux

We outline here the two most highly ranked: (1) expansion of the client’s commercial activity and (2) citizens purchasing goods and services locally.

Expansion of the client’s commercial activity: It cannot be conclusively proven that the improvement in clients’ living conditions through the development of their businesses is due solely to their participation in the CrediAmigo program. However, one can certainly argue that such participation played an important role. As a program manager explained:

I can’t guarantee that all of this increase in income was due to CrediAmigo. I can guarantee that CrediAmigo seems efficient, since the increase in income is very high, the increase in turnover is very high, and the exit from poverty of that client is growing. The speed at which that client exits poverty is 9% annually. The rate in other [organizations] such as BRAC in Bangladesh is 4-5%. So perhaps my credit is being efficient (...) but I can’t say that the microcredit has an impact of 9% of exit from poverty, because there are other factors included.

Gilberto, program manager

As the clients receive their loan, they can first increase their inventory. The credit allows them to pay cash for the merchandise, which gives them right to a discount. After a while, they can make renovations in their shop in order to enlarge it. Doing this, they can stock a greater variety of merchandise, which translates into more sales. In other cases, clients who were selling their merchandise on the street eventually can afford to have their own shop. Some of these people achieve this relying solely on the credit provided by the program.

(...) people who did not have a shop, who were streets vendors, now have their own shop. People who were selling broth at public events now have their own place, at home. (…) We can clearly see that, today, three years later.

Emmanuelle, credit agent

I was nothing! Today, I have a motorcycle (...) to get to deliver my orders [meals] faster. (...) This month, because we are at the end of the year, I had an increase (...) of 40%!

Joao, client

Citizens purchasing goods and services locally: At the beginning, many people had to go downtown to purchase goods and services essentially for two reasons. First, they could simply not find the goods they were looking for as there was not a great variety of shops in the neighbourhood. Second, few shop owners in the neighbourhood could afford to offer credit to their clients, which obliged many customers to purchase elsewhere. As more people become clients of CrediAmigo, and consequently succeed in expanding their business, the variety of goods and services offered within the neighbourhood increases. The provision of credit progressively allows more shop owners to finance their clients’ purchases. As these changes occur, an increasing number of people start to purchase more goods and services within their neighbourhood, accelerating the neighbourhood’s economic development. As a client explained:

When someone goes downtown: “Get this for me”. It is no longer necessary to go downtown. They know someone sells it in the neighbourhood and even offers the client to pay in multiple installments [as is the case downtown].

Margarida, client

This process was also facilitated by the fact that credit agents explain to the clients the importance of spending their money in the neighbourhood to further its development.

We credit agents, also have a social role, which is promoting economic development there. [We even encourage negotiation between people (...) promoting these people’s business, the purchase of merchandise between them... “If there is someone who can be a supplier, why would you go buy in Fortaleza if you have a friend here? Let’s go, let’s make a partnership with him”. We make a partnership, saying: “Look, this guy is my client, make a good price for all of my clients and I will talk about you and they will buy here”.

Leno, credit agent

With regard to the comparison by relevant social group, and to the literature review, we will stress four elements.

First, there seems to be a difference in the presentation of the consequences by the program managers, on the one hand, and the credit agents, on the other. The managers tended to be more careful when discussing the impact the program had on the clients and the neighbourhoods as a whole. They insisted that they could not tell what part of this economic development was due to CrediAmigo, considering that a variety of factors have certainly converged in some way. Still, they suggested that the program probably had an impact as the rate of development was considerably high. However, the credit agents did not hesitate to confirm that the consequences suggested were attributable to the program. In fact, during the interviews, we repeatedly noticed that the credit agents were generally very positive regarding the development of the program and its methodology, which they seemed to consider perfect. For instance, the lack of a critical stance can be perceived in the following comment, in which a credit agent discusses the consequences of CrediAmigo in the clients’ life:

We are modifying [read: renovating] their houses

Emmanuelle, credit agent

In this statement, the credit agent does not consider that the client does not necessarily count solely on CrediAmigo to finance the whole project. One also feels a strong identification with the program on the part of the credit agents. Consider, for instance, the following comment:

The consequences were the best we could expect, because when the bank created this product, it clearly developed the region, since it is one of its objectives to develop the Northeast and implement in other States. And it certainly succeeded. Many people benefited from that project. Perhaps if there had not been this project, today, they [clients] would be in the same situation as they were ten years ago; because there would not be any other way to be conceded an investment [loan] or training.

José, credit agent

A number of motives may explain this difference in stance between managers and credit agents, including: level of education, experience and position within the organization. The insistence on getting the credit agents to strongly internalize the methodology may also have influenced their perception.

Since credit agents are in the field most of the time, closer to the ultimate client, they tend to have a micro view instead of an overall view of the impact of credit, though the most efficient agents achieve to maintain a 300-400-client portfolio, a small fraction of BND operations. In addition, it is reasonable to assume that the return promoted by credit agent activities is more directly affected by credit performance. For this reason, the agent tends to have a strong incentive to emphasize, even exaggerate, the benefits of the extended loans.

Second, the addition of public services within the neighbourhood was not among the consequences suggested by clients. This omission may not be a coincidence, particularly considering that the clients comprise six respondents, the largest relevant social group. Now it is reasonable to think that the addition of public services (at least those most visible, such as infrastructure projects) takes place essentially in poorer and distant neighbourhoods, those that are most underserved by local governments. The CrediAmigo clients who live in these poorer areas usually receive a loan through the product Comunidade (communal banks), by which the program lends lower amounts to larger groups than is the case with solidarity groups. This being said, the clients we interviewed are part of solidarity groups and don’t live in those poorer areas. Therefore, they might not witness as many changes in provision of public services within their neighbourhood. As a result, they would not tend to bring up this issue during the interview. Should this conclusion be confirmed, the importance of the consequence discussed here should be put in perspective.

Third, the formalization of the client through provision of a bank account was not suggested as a consequence by the clients. On the one hand, it may be that the clients did not mention this consequence because the question focused on the development of their business and neighbourhood. Another explanation is that clients do not consider the provision of a bank account, say, as an important step in their life. Note that the respondents who mentioned this consequence as important did not suggest that clients perceived it as such.

Fourth, the increase in the number and variety of shops within the neighbourhood was only suggested as a consequence by the clients. We may think that the other groups did not mention it because they were not asked directly about the consequences for the neighbourhood. Following the same logic, the clients would have provided this answer because they had been asked this specific question and because this consequence may be the most evident change for a client to notice.


In this paper, we sought to increase our understanding of the conditions facilitating the replication of a given microcredit methodology – CrediAmigo – in a given context, that of Brazil. The analysis of the empirical material was guided by a theoretical framework integrating four main concepts, namely relevant social groups, interpretive frames, negotiation process and methodologies-in-practice. The results were compared to the literature review and they were also compared by relevant social group, as the perception of the program managers, coordinators, credit agents and clients regarding each of the questions asked could differ. Our findings corroborate that different social groups might have diverging and/or converging views on important issues related to the replication of CrediAmigo microcredit methodology and that taking these different perceptions and experiences into account is likely to improve the implementation process, increasing the replication process’ chances of success.

In terms of concrete practical contribution for microfinance practitioners, the results of our study have the potential to help managers of microcredit institutions or programs as well as Brazilian policy makers. We provide below four recommendations that may help to create conditions that facilitate the replication of microcredit methodologies from one context to another. Although they were developed from the investigation of a particular methodology – CrediAmigo – in a particular context – the Brazilian one, we strongly believe in the plausibility of generalizing these recommendations to other methodologies (mainly solidarity group-based methodologies) and other contexts, evidently with reflexivity and caution and open to future investigation.

In the case of the replication of CrediAmigo, focus on peer selection and enforcement at the initial stage of the implementation, and refocus on the credit agent’s role as the program matures, were perceived as two essential principles that contributed to the success of the implementation in the Northeast of Brazil, particularly in Fortaleza. These two principles could be important to the replication of other microcredit programs using the same underlying logic as CrediAmigo, i.e., the solidarity groups. Therefore,

Recommendation 1: We suggest that managers replicating microcredit programs based on solidarity group methodologies from one context to another should pay special attention to these two principles – peer selection and enforcement and commitment of the credit agent.

Because different social groups involved in the replication of a microcredit methodology are likely to have different perceptions of the most important principles to be taken into account during the process, it is important to be aware of these potential differences. Therefore,

Recommendation 2: We suggest that managers responsible for the replication of microcredit methodologies pay attention to different perceptions held by different social groups. Those perceptions may be divergent or conflicting. In that case, they should be addressed purposefully. They may be convergent or complementary. In this case, each group can bring different angles and experiences that, together, can help improve the implementation process and increase the chances of success of the replication process.

Because different conditions are likely to constitute barriers to the successful replication of a microcredit program, it is important to be clear on what these conditions are and the best way to cope with them in a given replication process. Doing this should increase the chances of achieving a successful replication. Among the conditions considered as barriers to replication in the literature, cultural conditions emerged as particularly important. Therefore,

Recommendation 3: We suggest that managers in charge of the replication of microcredit programs pay attention to these cultural conditions in order to better adapt to the locality where the program is being implemented.

Entering a highly competitive market such as Rio de Janeiro or Sao Paulo can be greatly facilitated by establishing a partnership deal with a local organization. This initiative can provide the organization implementing a microcredit program with initial contacts, a client portfolio, and a name recognized locally that strengthens the organization’s position in the market. In addition, seeking to include individuals from the local community in the organization’s implementation team may help to ease the program’s adaptation, furthering organizational learning and, consequently, flattening its organizational learning curve. Therefore,

Recommendation 4: We suggest that managers replicating microcredit methodologies in highly competitive markets search for local partnerships likely to increase their understanding of the local context, namely through tacit knowledge transfers.

These four recommendations represent the main contribution of our study for practice. From a theoretical perspective, the results of the analysis bring new light to the literature on methodology replication by confirming a number of assertions, suggesting new ones, and sometimes challenging current positions. To illustrate this, let us consider two themes discussed in the analysis and discussion of the results. First, among the perceived main principles of the CrediAmigo methodology, one confirmed the literature (“peer lending and enforcement”), whereas others (“commitment of the credit agent” and “management’s monitoring process”) could be considered complementary to existing literature. Second, of the conditions that facilitated the implementation of CrediAmigo in Fortaleza, some conditions were perceived as more significant than what the literature actually suggests (e.g., “cultural conditions”). These new considerations enrich the literature and may help managers and policy makers to come up with strategies and practices better suited to the provision of microfinance services in different regions and cities of Brazil.

Our paper also provides microfinance scholars with a fresh investigation of the replication of a Brazilian microcredit model, beyond the typical studies on international replication of, say, the Grameen Bank (Bangladesh) or Bank Rakyat (Indonesia). Interestingly, it shows how regional specificities may impact the success of a replication taking place at the national level, instead of internationally. Based on the results we obtained, we might think that before analyzing a phenomenon as complex as the international replication of a methodology, it is useful to investigate replication initiatives taking place within national boundaries.

Although our results are related to our specific research question, we consider them to have an important property of relevant research contributions – they are plausible. Such plausibility increases the possibility of these results being transferable to other contexts than the Brazilian one, as our four recommendations intend.