3. Methodology and scope
3.3. Data collection
3.3.2. Qualitative data
Qualitative data was collected through semi-structured interviews with different important stakeholders of the Latin American impact investment industry. The interviews were carried out in order to gain a more profound and updated understanding of the available quantitative data.
Moreover, additional, the interviews provided new information regarding the following topics:
1. Explanation of the dominance of foreign impact investors in Latin America
2. Strategies for impact investors to distinguish themselves from foreign impact investors
3. Industries that are particularly attractive for local impact investors
3.3.2.1. Semi structured interviews
The methodology of semi-structured interviews was selected for the collection of qualitative data. The primary reason for choosing semi-structured interviews was that this methodology allows to predetermine the content and to propose a structure of the interview on the one hand and to flexibly adapt the structure and the content to the topics addressed by the interviewee on the other hand (Longhurst, 2003). The semi-structured interviews contain questions and topics that must be covered but also allow the interviewer to additionally dive into issues that are brought up by the interviewee.
The interviews followed a prepared interview protocol (Appendix 21 and 22) (Creswell, 2014). The interview protocol, moreover, serves to provide reliability, consistency, and comparability regarding the information collected during the interviews.
3.3.2.2. Selection of the panel
A diverse interview panel served to combine and cross-check the insight generated through the interviews. The generated insights were, mover, triangulated with the quantitative data obtained from the industry databases. Thus, the combination of different points of view generates an as holistic image as possible. This diversification, further, helped to reduce the impact of individual biases on the consolidated findings. Main criteria for the identification of possible interviews were the following:
1. Professional and academic background: All interview candidates must be actively involved in the field of impact investing or social entrepreneurship. In order to generate different points of view it was important to select interviewees that play different roles in the sectors. Those roles include primarily:
a. Impact investors: Impact investors themselves are expected to know the impact investing industry best. Thus, interviews with insiders and industry experts were expected to yield the broadest insights.
b. Social entrepreneurs: Many social businesses represent attractive investment opportunities for impact investors. Consequently, social entrepreneurs that have dealt with or even received impact investors may provide additional when it comes to the intersect of social business and impact investing.
c. Other experts in the field of (social) entrepreneurship: Researchers, academics and professionals working for accelerators, incubators and other organizations related to (social) entrepreneurship and impact investing may provide additional, more generic information of the macro- and microenvironment of the social entrepreneurship and impact investing industry.
2. Regional differences: To gather not only quantitative but also qualitative information regarding both major Latin American markets, Mexico and Brazil, interviewees were selected from both countries. Comparing the additionally obtained information it was possible to identify common characterises of both markets and to pinpoint unique properties of the specific economy.
A process of three steps lead to the final interviewee sample. First, based on the beforementioned criteria a set of possible investors was identified. Second, citations and mentions in academic papers, industry reports and newspaper articles lead to the identified of the most recognized and important industry experts. Thus, they were expected to provide the most updated and deepest insights.
Therefore, they were contacted via e-mail and LinkedIn and asked for their availability willingness to carry out an interview. Third, the final set of interviewees resulted, accordingly, from the availability
of the interview partners and the convenience of executing the interviews. Thus, the combination of judgement, convenience and opportunity led to the final selection of the five interview partners Armando Laborde, Luis Fernando Laranja da Fonseca, Pedro Vilela, Enrico Robles and Angel Sahagun.
A detailed list of the identified, contacted and interviewed interview partners can be found in Appendix 32.
Following the sophistication method interviews were carried out until the quantitative data was deeply understood, triangulated and evaluated and responses repeated similar insights (Creswell & Poth, 2017) (Mason, 2010). Thus, a relatively smaller sample size of five interviewees generated enough additional insights. Each interview took between 30 minutes and an hour, so that time was not a limiting factor and interviewees could give detailed and high-quality insights. Thus, due to this quality of obtained information additional interviews with available interviews were not expected to yield new insights (Morse, 2000). One final reason for the sample size of five interviewees was the limited availability of identified interview partners as one natural limiting factor.
3.3.2.3. Technical execution
Due to the geographic separation, face-to-face interviews were not possible, and Skype- and telephone-interviews were carried out instead. The interviews were conducted either in English, Portuguese, or Spanish at the request of the interviewee. To facilitate the coding and analysis of all interviews, the records of the semi-structured were transcribed in their original language, using the professional transcription service AmberScript. All transcripts were checked manually to ensure the correctness. Transcripts in Portuguese and Spanish, were, in a second step, translated into English using Microsoft Office´s translation functionality. Again, to ensure correctness and consistency, all translations were checked manually. All English versions of the transcribed interview protocols were coded manually to ensure the comparability of the findings and to make efficient use of the gained insights. Due to technical problems, the interview with Angel Sahagun could not be recorded. The transcript of the interview with him is a memory protocol, written down immediately after the interview. All transcripts can be found in Appendix 24-31.