Over the past two decades, donor funding of international organizations (IOs) such as the United Nations, has shifted from core resources toward earmarked contributions. While core resources provide IOs with discretion how to achieve their mandates, earmarked contributions require IOs to focus their assistance on specific themes, sectors, countries, or projects, as determined by the donors. Our understanding of the drivers and effects of earmarked funding has been limited by the lack of systematic data. To address this gap, we introduce the Earmarked Funding Dataset. The dataset contains 342,812 earmarked aid activities from nearly 50 donors with over 340 IOs with a mandate in international development from 1990 to 2020. We first explore the data descriptively to cast light on the evolution of the multilateral development system. We then use the data to examine whether and how donors financially reward well-performing IOs. To that end, we combine our data with newly available performance data synthesized from Multilateral Organization Performance Assessment Network (MOPAN) evaluations of 32 IOs in 2009–2020. Using difference-in-difference estimations, we find a significantly positive relationship between outcome performance and earmarked contributions. In contrast, performance does not seem to affect core resources. These patterns suggest that donors reward well-performing multilaterals with more funding, albeit based on development outcomes which are not easily contractible. From a policy perspective, these findings suggest that performance-based approaches to aid allocation—as they are currently practiced—upload the financial risks of organizational reform processes to international organizations.

过去二十年来，联合国等国际组织（IO）的捐助资金已从核心资源转向专项捐款。虽然核心资源为国际组织提供了如何实现其任务的自由裁量权，但专用捐款要求国际组织将其援助重点放在捐助者确定的特定主题、部门、国家或项目上。由于缺乏系统数据，我们对专用资金的驱动因素和影响的理解受到限制。为了解决这一差距，我们引入了专用资金数据集。该数据集包含 1990 年至 2020 年间来自近 50 个捐助者和 340 多个担负国际发展任务的国际组织的 342,812 项指定援助活动。我们首先对数据进行描述性探索，以阐明多边发展体系的演变。然后，我们使用这些数据来检查捐助者是否以及如何对表现良好的国际组织进行经济奖励。为此，我们将我们的数据与多边组织绩效评估网络 (MOPAN) 对 2009 年至 2020 年 32 个国际组织的评估综合得出的最新绩效数据相结合。使用双重差分估计，我们发现结果表现和指定捐款之间存在显着的正相关关系。相比之下，性能似乎并不影响核心资源。这些模式表明，捐助者会向表现良好的多边组织提供更多资金，尽管其基础是不易收缩的发展成果。从政策角度来看，这些发现表明，目前实行的基于绩效的援助分配方法将组织改革进程的财务风险推给了国际组织。