Experimental Design Details
The primary data source for this experiment is data collected by the research team through the survey designed to examine the impact of Covid in Mozambican firms. As indicated before, this data is collected just minutes before the information provision that represents the core of our intervention. Initial information on firms was obtained from the National Statistics Institute containing data from the latest Enterprise Census (CEMPRE) performed in Mozambique from September 2015 to September 2016. We imposed three restrictions on the data: (i) firms were categorized by size using INE methodology, (ii) only micro and small firms were included and (iii) only firms that had available contact information. From these restrictions we had a population of 28677 firms.
The sampling strategy has two main goals:
1. To ensure the probability that treatment effects existed, diminishing the possibility that a firm could be interested in the policy presented but unwilling to show interest due to geographical distance.
2. Still ensure that the sample is representative across a set of baseline characteristics, namely, that firms in a more rural context – still an important part of Mozambican economic fabric – were covered.
Another important aspect of our sampling strategy is that the present study is part of an ongoing information experiment designed together with key stakeholders of Mozambique, namely The Ministry of Economics and Finance (MEF) and the Institute for the Promotion of Small and Medium Enterprises (IPEME). As part of the initiatives promoted by IPEME, the Institute has 9 Entrepreneur Guidance Centres (EGC) distributed across 6 provinces. In order to have a representative sample of the country while accounting for the importance of the EGCs in our research, 75% of the firms that our data collection team tried to reach were located within 25 km of an EGC. The remaining 25% were selected from companies located outside this radius.
Our team used Computer Assisted Telephone Interviewing (CATI) through the Kobo Toolbox data collecting platform.
To validate participation, all firm managers had to confirm some basic data information, such as the name and the region of the firm. Moreover, all participants were informed about the purpose of the present study and consented to share information. They had the option to interrupt the interview at any point and to refuse or skip any question according to their desire.
Our survey aimed at providing critical information to help policy makers monitor the effects of the pandemic on businesses. Therefore, the questionnaire was divided into 5 different key sections: firms’ main characteristics, business activity, covid-19, public policies, and managers’ expectations.
All firms participating in the study were randomly assigned to either a control group, a “consulting” treatment group (T1) or a “training” treatment group (T2). Each of these treatments consists in the enumerator reading to the interviewee a small text detailing information about existing public services that provide support to his/her firm. Enumerators read a small general text to participants in the control group detailing the intent of the government to provide public support to firms. The first treatment focuses on promoting public enterprise consulting services, the second treatment promotes entrepreneurial training and the provision of relevant information. The two treatments specify that the government offers these services through IPEME, the Mozambican public agency tasked with promoting and developing MSMEs in the country. Both texts emphasise the ways in which businesses could benefit from the corresponding services. After this information n is provided to the manager, we ask several questions, including: (i) preferred sub-services preferred among those offered (Legal; HR; Accounting; Management and Production), (ii) the willingness to pay for these preferred services and (iii) whether the entrepreneur would prefer these services to be provided publicly or privately. Finally, respondents are asked to provide their phone number if they are interested in being contacted by our research team with further information about IPEME services. Two SMS are then sent to those participants who expressed their interest in receiving further information. The first is sent on the same day as the call, the second SMS will be sent a few days after the end of the call and finally an e-mail will be sent. These SMSs ask participants to respond to the message if they would like to receive more information about the category of services specific to their treatment (or control) group.