A carregar...
Publicação
Rural development projects in Latin America: understanding the critical factors for sustained impact
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.9 MB | Adobe PDF |
Orientador(es)
Resumo(s)
Nas últimas décadas, a América Latina tem recebido níveis substanciais de ajuda externa para o desenvolvimento rural. No entanto, as comunidades não registaram melhorias significativas nos seus meios de subsistência e continuam a suportar uma pobreza persistente. Embora as iniciativas atinjam frequentemente os seus objectivos imediatos ao nível dos indicadores, é menos frequente que atinjam os resultados pretendidos e raramente produzem um impacto sustentável a longo prazo. Com o objetivo de compreender os factores críticos para intervenções bem sucedidas, esta tese começa por apresentar uma abordagem integrada que reúne três perspectivas fundamentais, proporcionando uma análise abrangente adequada à complexidade do problema de investigação. Uma vez que a compreensão do contexto em que os projectos são implementados é fundamental para alcançar os resultados esperados, é desenvolvido um quadro específico para captar os principais aspectos contextuais, estruturado em torno de quatro dimensões, 14 temas e 33 parâmetros. Utilizando este quadro, são utilizadas técnicas de aprendizagem automática para agrupar países com base em caraterísticas partilhadas. São formados oito grupos, alguns dos quais apresentam coerência geográfica, como os países insulares das Caraíbas ou os do norte da América Central. Outros clusters, como os que agrupam o Brasil com o Equador e o Panamá, revelam semelhanças menos intuitivas. Posteriormente, as caraterísticas intrínsecas dos projectos de desenvolvimento são examinadas através da lente deste quadro para compreender como e em que medida os princípios incorporados influenciam a realização dos objectivos desejados. Centrando-se em 50 anos de iniciativas de desenvolvimento rural na Guatemala, os resultados mostram que a inclusão e a responsabilização são os factores mais críticos para o êxito dos resultados finais. Por último, a interligação entre as estratégias de implementação e os resultados pós-intervenção é examinada através de uma combinação de técnicas de Processamento de Linguagem Natural e investigação participativa realizada durante o trabalho de campo. Esta perspetiva destaca a capacitação, a harmonização e as parcerias como os princípios mais importantes para permitir a sustentabilidade a longo prazo dos resultados do projeto. Esta investigação fornece provas de que os resultados e a sustentabilidade das intervenções de desenvolvimento rural podem ser melhorados através da adoção dos factores críticos identificados ao longo do ciclo do projeto, desde a conceção até à implementação. Embora os projectos na Guatemala sejam investigados como estudo de caso, as conclusões são amplamente aplicáveis a outros países com caraterísticas semelhantes, tal como identificadas através do processo de agrupamento.
Over the past decades, Latin America has received substantial levels of foreign aid for rural development. However, communities have not consistently experienced significant improvements in livelihoods and continue to endure persistent poverty. While initiatives often achieve their immediate output-level objectives, they less frequently realize their intended outcomes and rarely deliver long-term, sustainable impact. Aiming at understanding critical factors for successful interventions, this thesis starts by presenting an integrated approach that bridges three key perspectives, providing a comprehensive analysis suited to the complexity of the research problem. Since understanding the context in which projects are implemented is critical for achieving their expected outcomes, a dedicated framework is developed to capture key contextual aspects, structured around four dimensions, 14 themes and 33 parameters. Using this framework, Machine Learning Techniques are employed to cluster countries based on shared characteristics. Eight clusters are formed, some exhibiting geographical coherence, such as the island countries in the Caribbean or those in northern Central America. Other clusters, such as those grouping Brazil with Ecuador and Panama, reveal less intuitive similarities. Subsequently, the intrinsic characteristics of development projects are examined through the lens of this framework to understand how and to what extent embedded principles influence the achievement of desired goals. Focusing on 50 years of rural development initiatives in Guatemala, results show that inclusion and accountability are the most critical factors for successful end-line results. Lastly, the interconnectedness between implementation strategies and post-intervention outcomes is examined through a combination of Natural Language Processing techniques and participatory research conducted during fieldwork. This perspective highlights empowerment, harmonization and partnerships as most pivotal principles to enable long-term sustainability of project results. This research provides evidence that results and sustainability of rural development interventions can be enhanced by adopting the identified critical factors throughout the project cycle, from design to implementation. Although projects in Guatemala are researched as case-study, the findings are broadly applicable to other countries with similar characteristics, as identified through the clustering process.
Over the past decades, Latin America has received substantial levels of foreign aid for rural development. However, communities have not consistently experienced significant improvements in livelihoods and continue to endure persistent poverty. While initiatives often achieve their immediate output-level objectives, they less frequently realize their intended outcomes and rarely deliver long-term, sustainable impact. Aiming at understanding critical factors for successful interventions, this thesis starts by presenting an integrated approach that bridges three key perspectives, providing a comprehensive analysis suited to the complexity of the research problem. Since understanding the context in which projects are implemented is critical for achieving their expected outcomes, a dedicated framework is developed to capture key contextual aspects, structured around four dimensions, 14 themes and 33 parameters. Using this framework, Machine Learning Techniques are employed to cluster countries based on shared characteristics. Eight clusters are formed, some exhibiting geographical coherence, such as the island countries in the Caribbean or those in northern Central America. Other clusters, such as those grouping Brazil with Ecuador and Panama, reveal less intuitive similarities. Subsequently, the intrinsic characteristics of development projects are examined through the lens of this framework to understand how and to what extent embedded principles influence the achievement of desired goals. Focusing on 50 years of rural development initiatives in Guatemala, results show that inclusion and accountability are the most critical factors for successful end-line results. Lastly, the interconnectedness between implementation strategies and post-intervention outcomes is examined through a combination of Natural Language Processing techniques and participatory research conducted during fieldwork. This perspective highlights empowerment, harmonization and partnerships as most pivotal principles to enable long-term sustainability of project results. This research provides evidence that results and sustainability of rural development interventions can be enhanced by adopting the identified critical factors throughout the project cycle, from design to implementation. Although projects in Guatemala are researched as case-study, the findings are broadly applicable to other countries with similar characteristics, as identified through the clustering process.
Descrição
Tese de doutoramento, Estudos do Desenvolvimento, 2025, Universidade de Lisboa, Instituto Superior de Agronomia
Palavras-chave
Latin America rural development project success project sustainability machine learning América Latina desenvolvimento rural sucesso do projeto sustentabilidade do projeto aprendizagem automática