In Dar es Salaam, Tanzania (DSM) a pre-test post-test intervention pilot study with four arms was conducted. Each arm had 12-15 participants who were young men at risk for violence who were members of one 'camp'. Each participant was interviewed at baseline before the 4 camps were randomized.

The interventions were training sessions in: Health (Control), Entrepreneurship, Beekeeping, and All Intervention sessions. Subsequent interviews took place 3 months, 6 months, and 1 year after all the interventions took place.

This was a pilot study in preparation for a full intervention trial. The hypotheses for the intervention trial were expected to be that income would increase and violence would decrease in the intervention arms as compared to the Control arm.

Most violence in Dar es Salaam Tanzania is committed by and against young men. The most significant predictors of homicide death are unemployment, poor education, and living alone. Young, under-employed men with poor education are also at risk of recruitment into non-national armies such as al-Shabaab. However, they have the potential to contribute to be a positive force behind national economic empowerment with the right training and support.

Intervention around employment status is the most achievable. Most occupations open to uneducated young men are inherently unstable as they are generally seasonal, short term, or illegal. With formal jobs in short supply, difficulty in generating more, and a large youth population to absorb, most governments in Africa and multinational financial institutions agree that entrepreneurship is an important option to securing economic and political security.

Self-employment and entrepreneurship is encouraged by the Tanzanian government. Several initiatives have emerged in Tanzania to address youth unemployment. Non-governmental organizations (NGOs) including International Labour Organization (ILO), Plan International, Restless Development and others have started implementing projects to address the lack of opportunities for productive economic engagement by young men. Novel strategies include establishment of bee-keeping enterprises in youth groups supported by the Tanzania Forest Services (TFS) at the Ministry of Natural Resources and Tourism, and infotainment to give youth skills enabling entrepreneurship developed by Fema, an NGO. Little research data are available on the effectiveness of such initiatives even though they have good potential for positive results.

Few studies have examined entrepreneurial and micro-enterprise programs with a focus on young men at risk for violence, especially from Africa. Most research on entrepreneurial and micro-enterprise programs in low-income countries center around the issue of HIV/AIDS with a focus on women who have not been the traditional income earners. In addition to a focus on a different population, interventions are often poorly described, and programmatic content was routinely omitted. Studies are usually a single group post-intervention assessment using quantitative, qualitative or mixed methods. Most are also flawed due to small sample sizes, lack of controls, or objective longitudinal measures. Thus there is a dearth of rigorous research on entrepreneurial and micro-enterprise programs.

In order to address this research gap, an intervention trial with four arms involving training in entrepreneurship, beekeeping, and health education was envisioned.

Pilot studies are an important step to any major trial which seeks to evaluate an intervention. Here the investigators report on a pilot study focusing on entrepreneurship, beekeeping, and health training conducted to inform such an intervention trial. The aims for this pilot study included:

1. Evaluating the suitability of the camps as places to recruit the target group
2. Evaluating the suitability of the curricula.
3. Quantifying bee product production after one year
4. Gaining insight into income generation potential of four interventions after 3, 6, and 12 months
5. Obtaining the required preliminary data for the calculation of a sample size for the primary outcome

Methods

Site and setting The site of the study was Dar es Salaam (DSM), the largest city in Tanzania and a commercial hub for most of East and Central Africa. The most recent population census places DSM's population at an estimated 4.4 million. Young adults (ages 15-35 years) make up about 46.8% of the total population in Dar es Salaam Region, the larger jurisdiction of which the city of DSM is a part.

In Tanzania, young men often congregate in camps. These camps have names, fixed memberships, structured leadership, and specific meeting places; an individual's membership can last for years. Though not formally registered, camps are informally recognized by local leaders in the city. There are lots of camps generally scattered in all districts of DSM Region. The camps that were the focus of the present study were based in Kinondoni District, the largest of the five districts in the city. Detailed characteristics of the camps and camp members have been described; basically, the members are usually young men who are inadequately educated and under- or unemployed, and therefore at risk for stealing and being injured or killed in reprisal.

Sampling procedure

A total of 71 camps had previously been mapped and described. In these camps, six important variables were identified: camp name, number of male and female members, if the presence of a weapon was reported, if the interviewer felt safe, and whether the interviewer was told not to work there. Camps at which a weapon was reported (n = 20) or at which previous researchers had been asked not to work for safety reasons (n = 5), or where the interviewer did not feel safe (n = 9), were not included. Camps that had been chosen for previous research (n = 23) were also excluded. That left us with 14 camps from which to select our study population. Among these, 8 were not chosen because members were greater than 15 in number or because the camp had many female members. The investigators visited the remaining six camps to discuss the project with the camp leaders. Two of the camps were not selected because their members were not interested in beekeeping. Therefore, four camps remained for the project. Then the researchers introduced themselves to the local government leaders of the areas in which those four camps existed.

Data collection

Using the same questionnaire, respondents were interviewed four times: at baseline, and then at 3 months, 6 months, and one-year post-intervention.The same team of 10 interviewers conducted all interviews. Respondents were invited to Muhimbili University of Health and Allied Sciences (MUHAS) campus for face-to-face interviews. Respondents who could not come were contacted by telephone; if willing to be interviewed, they were interviewed at their camp site. Since several questions were potentially sensitive, respondents were informed that they should refuse any question about which they felt uncomfortable, rather than give an untrue answer.

Variables

Collected variables included age, education, occupation, number of dependents, an asset index, and exposures to violence. Based on the collected data from the open-ended question, occupation was categorized into five categories: micro-business; drivers; unskilled and day laborers; semi-skilled workers having gone through at least a minimum of specialized training; and those who were dependent on family members, including full-time students. The asset index was created from ownership of bicycle, motorcycle, cell phone, bed, plot, housing, floor, and toilet.Weekly income was self-reported.

Analysis

Quantitative The proportion of participants were compared across the intervention arms using chi-squared tests or exact tests, as appropriate. Differences in weekly reported income were examined at baseline between groups using t-tests and graphical analysis with the Control arm serving as the reference arm. Differences in average weekly income between baseline and each time points were also assessed and tested using t-tests and across groups using analysis of variance. A linear trend in income was investigated by regressing income as a continuous variable on-time point, with the correlation between multiple measures per person accounted for using generalized estimating equations to correct standard errors. Individual patterns of income change over the duration of the study were examined using spaghetti plots.

Qualitative

At the time of the final interview, Focus Group Discussions (FGD) were held. An FGD guide was used by the moderator to steer the discussions, with the following questions:

What did the participants like the most about the entrepreneurship, beekeeping, and health project? What expectations did the participants have? Were the participants' expectations met? Did their life change? How could the project be improved?

Inclusion Criteria:

• poorly educated, under- or un-employed, Tanzanian

Exclusion Criteria:

-

male

The study population was under or unemployed (less than minimum wage), poorly educated young men in Dar es Salaam, Tanzania.

All of the individual participant data collected during the trial, after de-identification will be available including the study protocol, statistical analysis plan, informed consent to anyone who wishes to access the data, for any purpose, indefinitely.

The data are available now and indefinitely.

To anybody with a request.