Do Safety Expectations Affect the Location Strategies of Large Service Delivery INGOs?

Abstract

Large service-delivery international nongovernmental organizations (INGOs) help shape the lives of millions worldwide, contributing significantly to the betterment of global health, development, education, and more. However, these organizations operate on limited financial resources and must deliver services on-the-ground, requiring them to make calculated decisions about their location strategies for operations. Although these INGOs should aim to maximize their efforts by operating in countries with the greatest need, sometimes they may face challenges from political conditions in these environments. Do safety expectations in countries affect the assistance allocation decisions of large service delivery INGOs? We explore this question using new data on the location and operational-intensity of country-level projects across 90 of the largest service delivery INGOs worldwide between 1990 and 2015. We find that these INGOs do send greater operational resources to countries demonstrating need and that INGOs are not necessarily afraid to operate in countries with safety risks. However, large INGOs do send far fewer service operations to autocracies, suggesting that political factors affect their location strategies as well.

Las grandes organizaciones no gubernamentales internacionales (ONGI) de prestación de servicios ayudan a dar forma a la vida de millones de personas en todo el mundo, contribuyendo significativamente a la mejoría de la salud mundial, al desarrollo y a la educación, entre otras muchas cosas. Sin embargo, estas organizaciones operan con recursos financieros limitados y tienen que prestar sus servicios sobre el terreno, lo que les obliga a tomar decisiones calculadas sobre sus estrategias de ubicación para las operaciones. Aunque, en teoría, estas ONGI deberían tratar de maximizar sus esfuerzos operando en los países que tienen unas mayores necesidades, a veces las ONGI pueden tener que enfrentarse a ciertos desafíos debido a las condiciones políticas existentes en estos entornos. ¿Afectan las expectativas de seguridad de los países a las decisiones de asignación de asistencia por parte de las grandes ONGI de prestación de servicios? Estudiamos esta cuestión utilizando nuevos datos acerca de la ubicación y de la intensidad operativa de los proyectos a nivel de país en las 90 ONGI de prestación de servicios más grandes del mundo entre 1990 y 2015. Encontramos que estas ONGI envían mayores recursos operativos a los países que demuestran una mayor necesidad y que las ONGI no necesariamente tienen miedo a operar en países con riesgos de seguridad. Sin embargo, las grandes ONGI envían muchas menos operaciones de servicio a las autocracias, lo que sugiere que los factores políticos también afectan sus estrategias de localización.

Les grandes organisations non gouvernementales internationales (ONGI) qui fournissent des services ont une incidence sur la vie de millions de personnes dans le monde entier, notamment en contribuant considérablement à l’amélioration de la santé, du développement et de l’éducation à l’échelle mondiale. Néanmoins, ces organisations fonctionnent à partir de ressources financières limitées et doivent fournir des services sur le terrain. Elles doivent donc prendre des décisions réfléchies quant à leur stratégie d’emplacement pour leurs opérations. Bien que ces ONGI devraient souhaiter optimiser leurs efforts en concentrant leurs activités dans les pays où les besoins sont les plus importants, elles peuvent parfois se retrouver confrontées à des défis issus des conditions politiques de ces environnements. Les attentes en matière de sécurité dans les pays ont-elles une incidence sur les décisions d’attribution d’aides des grandes ONGI qui fournissent des services ? Nous examinons cette question à l’aide de nouvelles données sur l’emplacement et l’intensité des activités dans les projets au niveau national chez 90 des plus grandes ONGI qui fournissent des services dans le monde entier entre 1990 et 2015. Nous observons que ces ONGI envoient bel et bien plus de ressources opérationnelles dans les pays qui démontrent un besoin et qu’elles n’ont pas particulièrement peur d’aller dans les pays où il existe des menaces pour la sécurité. Néanmoins, les grandes ONGI lancent bien moins d’opérations de fourniture de services dans les autocraties, ce qui suggère que les facteurs politiques influencent aussi leurs stratégies de choix d’emplacement.

Keywords:

• Quantitative analysis
• service delivery NGOs

Notes

1 See the supplemental appendix for specific qualifiers used to determine “large” and “service delivery” in our sample of INGOs. Replication data available through International Interactions Harvard Dataverse: dataverse.harvard.edu/dataverse/internationalinteractions.

2 INGO budgets have blossomed—the International Committee of the Red Cross (ICRC) reported around $1.9 billion USD in donations in 2020, financing 18,827 staff in more than one hundred countries (International Committee of the Red Cross 2020). Amnesty International had global expenditures of $325.3 million USD (Amnesty International 2020).

3 The temporal span aimed to cover the post-Cold War period, with data collection taking place in 2015.

4 Although many organizations met these criteria, there were several near the cut-off point that we determined were close enough to merit inclusion, based on firmly qualifying in other areas.

5 Other works aiming to quantify the number of INGOs or the scope of their operations have relied on the Yearbook of International Organizations (2020), such as Boli and Thomas (1999). These volumes provide profiles of INGOs, which include a range of descriptive information on their structures, missions, physical presences, and histories. However, we only utilized YIO as a lower-tier confirmatory source in our data. We were hesitant to rely on the YIO as a primary source, as opposed to annual reports, because there are issues with data construction that are inherent to the volumes’ publication method. The yearbook is self-reported, meaning the information included is proffered by the organizations each year. When INGOs do not report changes or simply do not report to the YIO in a particular year, information is often duplicated from the last reported year. For our purposes, this means that the countries of operation may remain unchanged (erroneously) for a spell of years, which would negatively affect our data’s structure and purpose. See Murdie (2014) for another version of these critiques of the YIO’s reporting methodology and how it affects data collection. To assess differences between our data and YIO’s more limited sample, we performed a correlation matrix using the somewhat larger interpolated YIO “delivery INGO” count variable from Campbell, DiGiuseppe, and Murdie (2019). We find moderate positive correlation between their variable and our three count transformations, demonstrating substantive differences, as seen in Table A10 of the supplemental appendix.

6 For example, annual budget expenditures are difficult to gather globally across a 25-year period covering 90 organizations.

7 We choose to use country counts of INGO activity because we were unable to gather more detailed and longitudinal data on data that better captured the magnitude of involvement, such as annual budgets.

8 For more on estimating non-negative dependent variables (with zero count varying) using Poisson pseudo-maximum likelihood (PPML), see Santos Silva and Tenreyro (2011).