Examining the effectiveness of four intervention types across five countries

Author: Joy Muthanje Mwaniki

Review: Moritz Stumpe

Thanks to Chris Bryant for providing feedback on our write up, and to Natalie Lehr for her editing contributions. All opinions and errors are our own. We highly welcome feedback and encourage the idea of red-teaming or challenging the conclusions made in our report.

Introduction

This paper analyses and compares different forms of animal advocacy interventions targeting farm or production animals across five African countries representing each region of the African continent: Egypt, Ghana, Kenya, Rwanda, and South Africa. The purpose of this study is to determine the interventions that are likely to succeed in these countries and inform animal welfare practitioners and organisations which interventions would be well suited to each country's context. Egypt, Kenya, and South Africa are included in this study due to their ranking on the Animal Protection Index (API), which rates countries based on their policies and legislation on animal welfare and provides helpful information for researchers (Owino, 2020, World Animal Protection, n.d.). Ghana and Rwanda are not included in the API but are selected as representative countries of their region to enhance the limited information available on their animal welfare ecosystems.

Methodology

We use a Weighted Factor Model (WFM) framework to analyse four types of interventions: 

• individual or public outreach, which targets behavioural change among the public;

• institutional outreach, which seeks to influence political institutions and corporations;

• capacity building, which focuses on improving the capacities of stakeholders in animal welfare; and 

• direct help, which provides direct services to animals, such as veterinary services, vaccinations, and animal rescues.

We use this framework because it is useful in situations where a large number of different quantitative and qualitative factors need to be consolidated, as is the case for this study. It also helps in identifying factors that decisively influence the final scores and thus supports clear communication and understanding (Charity Entrepreneurship, n.d.). The criteria, metrics, and weights used in the model are shown in Table 1 and are determined through our previous research, which can be found here. The criteria are weighted subjectively based on their estimated importance for impact, and the likelihood that each criterion will result in the failure of each intervention. By researching and gathering evidence on the four intervention types, they are scored in each country on a scale of one to five for each metric, with one always representing the least and five the most favourable score.

Table 1: Criteria, Weights and Metrics for Weighted Factor Model

Results

The final scores resulting from the Weighted Factor Model can be found in Table 2 below:

In all five countries, the most impactful type of intervention is individual or public outreach followed by institutional outreach. These similarities in scoring across countries are likely caused by the fact that the countries share commonalities in terms of socio-economic factors and culture. In South Africa, institutional outreach scores very highly, almost on par with individual outreach. Direct help does not seem promising at all in South Africa, while it seems relatively more promising in Egypt and Kenya than in the other countries. In Rwanda, individual outreach clearly outranks the other possible intervention types.

To test the robustness of the model, different sensitivity analyses are conducted on the most relevant results. This is done by varying the weights of the different metrics on a range of +/- 2 percentage points and the scores by +/- 0.5.  This method shows that the rankings of individual and institutional outreach in South Africa are robust, with individual outreach outperforming institutional outreach in 88% of the simulated cases. This is true even though the difference between the two scores is very small – less than 3% in 56% of the cases. As South Africa scores highest on institutional outreach, individual outreach is even more likely to outscore institutional outreach in all of the other four countries. A similar analysis also finds that Ghana or Kenya is very likely to have the highest scores for individual outreach among the five countries. In 83% of the simulated cases, Ghana achieves higher scores than South Africa on individual outreach, even though the difference between the scores of two countries is small – less than 3% in 69% of the cases. Since these analyses focus on the more contested results of the model (where differences in scores are not very large), other findings should be even more robust, such as the result that direct help in South Africa should be the least promising intervention. For interested readers, the detailed sensitivity analyses can be found here.

The scores for individual outreach are generally highest because this intervention does not demonstrate major weaknesses on the metrics covered in the WFM. Compared to other interventions, it is highly scalable, has a large talent pool to draw from and has good timing. In contrast, institutional outreach suffers from a high risk of no or negative impact across all of the five countries, even though it typically scores high on cost-effectiveness and neglectedness. Capacity building generally performs weaker across most metrics, most crucially on metric focus, as the impact of capacity building efforts are hard to measure. Direct help scores weakly across most of the metrics, even though it has a low risk of no or negative impact and tackles a significant problem in the five countries studied.

Practical implications

To give an example of the evidence and reasoning applied across this report, consider the high risk of no or negative impact for institutional outreach. In South Africa, political and corporate institutions may ignore lobbying efforts like petitions and protests (Fairbrother, 2017; National Council of SPCAs, n.d.). Such a risk is also present in Kenya where there are additionally no documented prosecutions for animal abuse (ICPALD, 2017). Similarly, in Egypt, there is a high risk of no impact as legislation often results in no change in the situation for animals (Salah, 2017, World Animal Protection, 2020). Further, in Rwanda, there is evidence that institutional interventions run the risk of resulting in errors in legislative reforms (Sherman, et al., 2014). Lastly, in Ghana, social conflict seems likely if legislation were to be changed to require the stunning of animals before slaughter (Agentur für Wirtschaft & Entwicklung, n.d.). More insights and the exact reasoning for every score can be found in the full report, especially in the results section.

Individual outreach is likely to be most impactful in all five countries, which suggests that animal advocacy organisations and individuals should first aim to create awareness among the public. Campaigns could be focused on animal welfare, sentience, or diet change. When designing such campaigns, advocates should keep in mind that there are risks of negative reactions from the public (such as perceptions that animal welfare and diets are un-African) and conflict with local cultural traditions. It is also noteworthy that within the Effective Altruism community, institutional outreach is often preferred over individual outreach (Reese Anthis, 2020), in contrast to the findings of this report. Nevertheless, prioritising public outreach efforts in the African countries studied here seems very reasonable, given their peculiar local context which has not been analysed thoroughly before. Public outreach may not be as effective as institutional outreach in places where awareness of animal rights and welfare issues is already quite high (such as the United Kingdom or United States). However, public outreach can be very relevant in communities where the animal advocacy movement is still small (Harris, 2020).

Limitations

That being said, several limitations concerning this study should be noted. Most importantly, there is limited information concerning animal welfare interventions across the African continent. This is especially true in the case of Egypt where missing information has been filled in by using information from comparable countries. In addition, information on the cost-effectiveness of animal welfare interventions in the five countries is notably absent, except for the cost-effectiveness of direct help. Again, information has been substituted and drawn from other contexts. Furthermore, this Weighted Factor Model is an initial try at scoring and is based purely on desk research, not on practical experiments. Further primary qualitative and quantitative data is required to draw more robust conclusions. More specifically, the use of qualitative interviews with experts from the relevant countries could likely fill in some of the gaps in the study, especially in the case of Egypt.