Philanthropy for global mental health 2000–2015

Background Mental disorders are the leading cause of years lived with disability worldwide. While over three-quarters of people with mental disorders live in low- and middle-income countries (LMICs) and effective low-cost interventions are available, resource commitments are extremely limited. This paper seeks to understand the role of philanthropy in this area and to inform discussions about how to increase investments. Methods Novel analyses of a dataset on development assistance for health were conducted to study philanthropic development assistance for mental health (DAMH) in 156 countries between 2000 and 2015. Results Philanthropic contributions more than doubled over 16 years, accounting for one-third (US$364.1 million) of total DAMH 2000–2015. However, across health conditions, mental disorders received the lowest amount of philanthropic development assistance for health (0.5%). Thirty-seven of 156 LMICs received no philanthropic DAMH between 2000 and 2015 and just three LMICs (Antigua and Barbuda, Grenada, Saint Vincent and the Grenadines) received more than US$1 philanthropic DAMH per capita over the entire period. Eighty-one percent of philanthropic DAMH was disbursed to unspecified locations. Conclusions Philanthropic donors are potentially playing a critical role in DAMH, and the paper identifies challenges and opportunities for increasing their impact in sustainable financing for mental health.


Appendix 1. Data sources and analyses
Data sources I merged the Institute of Health Metrics and Evaluation (IHME) dataset on development assistance for health (DAH) 1990-2017(IHME, 2018 with three variables: country classification per region (WHO, 2018), per country income-level (World Bank, 2018), and country population size (Global Burden of Disease Collaborative Network, 2018). DAH includes "in-kind and financial resources transferred from primary development channels to low-income and middle-income countries for the purpose of maintaining or improving health" (Dieleman et al., 2016(Dieleman et al., , p. 2537. The IHME DAH dataset reports semi-aggregated data on DAH in 172 countries between 1990 and 2017 (IHME, 2018). It reports estimates on resource flows from funding sources (Table 1), through channel organisations, defined as intermediary organisations disbursing funding to implementing institutions providing support in low-and middle-income countries ( Table 2). The dataset is built by IHME using different sources: Development Assistance Committee and Creditor Reporting System databases (Organisation for Economic Cooperation and Development), financial reports, audited financial statements, United States Agency for International Development Report of Voluntary Agencies, Foundation Center's grant database, Bill & Melinda Gates Foundation online grant database, Internal Revenue Service 990 tax forms, and personal correspondences (Global Burden of Disease Health Financing Collaborator Network, 2018).
These data, in an aggregated form, exists publically on the Global Health Data Exchange (IHME, 2018). A detailed dataset was obtained from IHME in September 2018, including values omitted in the publicly available dataset (i.e. values greater than US$0 but less than US$500, or less than US$0 and greater than -US$500). In addition, disaggregated data for United States foundations (variable channel, category Other US Foundations in Table 2) were obtained in June 2018.
It is worth noting that development assistance for mental health in the IHME DAH dataset captures not only mental disorders (including substance use disorders, dementia, and selfharm) but also some neurological conditions (epilepsy, headache disorders, Parkinson's disease). This reflects previous conceptualisations of mental disorders (WHO, 2008). At the time of the analyses for this paper it was not possible to access data on development assistance for mental health excluding those neurological conditions.

Analyses
I conducted descriptive analyses of annual philanthropic development assistance for mental health (DAMH) in absolute and relative terms, by channel organisation, by recipient country, and compared with philanthropic DAH to other health conditions (HIV/AIDS, tuberculosis, malaria, other infectious diseases, maternal health, newborn and child health, noncommunicable diseases excluding mental health). Philanthropic donors included in the analyses are corporations, foundations, individuals (Table 1-2). Analyses were limited to 2000-2015, due to poor data quality pre-2000, preliminary estimates post-2015 and to focus on the Millennium Development Goals era to inform the Sustainable Development Goals (SDGs) era, leaving 168 countries.
I excluded 12 small overseas territories or dependencies due to lack of World Bank country classification: Anguilla, Cook Islands, Mayotte, Montserrat, Nauru, Niue, Saint Helena, Saint Martin, Tokelau, Turks and Caicos Islands, Tuvalu, Wallis and Futuna Islands. None of them received philanthropic DAMH. Only two countries received non-philanthropic DAMH during the period, Anguilla (2005) and the Cook Islands (2005-2006 and 2008-2012). To reflect disbursements to recipient countries dissolved or created during the period of study (Kosovo, Serbia, South Sudan), the World Bank country classification was imputed using the first observation carried backward and the last observation carried forward.
Transfers between channels captured elsewhere in the database were excluded to avoid double-counting. Values are reported in 2017 United States dollars (US$) adjusted by purchasing-power parity. Analyses were conducted in Stata 14.