It was merely a few months ago that I swapped academic research on biological systems to research international data and metadata standards and how they relate to one another. Natural sciences have a special place in my heart so it is fitting that my first blog post marries the two fields.
My first task involved ‘sectors’ – functional classifications. There are loads of standards with loads of acronyms – CRS, COFOG, ISIC, NTEE to name but a few – all describing the same things in different ways. I needed to cross-map these standards to see what joins up, and what doesn’t. I started by studying the history of the international organisations responsible for these standards, discovering the origins of the classifications, how they were created and how prevalent these given classifications are now. Why, for instance do the Organisation for Economic Co-operation and Development (OECD) Development Assistance Committee (DAC) sector codes bear no relationship to the UN-maintained Classification of the Functions of Government (COFOG) standard, also developed by the OECD?
Having started to track the ‘how’ and the ‘who’, I am now eager to explore the ‘so what?’ factor: can joining up data standards help add a useful perspective to our understanding of the real world? With this in mind, I looked in detail at malaria in Uganda.
Malaria is endemic in 95% of Uganda and the climate promotes year-round transmission with little variability. The economic burden of the disease is heavy: it has been reported to consume up to 40% of African states’ health budgets, cause up to a 1.3% reduction in growth and losses of up to US$12 billion (USAID 2009). According to the World Health Organization (WHO), malaria significantly limits the economic and social development desperately needed to increase gross domestic product (GDP) rates.
The new Sustainable Development Goals (SDGs) include a target to, by 2030, “end the epidemics of AIDS, tuberculosis, malaria and neglected tropical diseases and combat hepatitis, water-borne diseases and other communicable diseases”. Given the figures from USAID and the WHO above, achieving this goal is clearly going to be a challenge.
In my research, I came across a recently published paper by Barofsky et al, 2015 that looks at the long-term impact of malaria eradication in south western Uganda and the positive effects it has on educational attainment and economic performance (when compared with the rest of Uganda). Barofsky argues that these benefits allow planners to reduce the imputed costs of eradication programmes, but it is critical that international and domestic resources are coordinated to maximise both impact and cost benefit.
This raises the interesting question for me, as a data scientist working on interoperability, of whether data on international aid flows and data on local-level health spending can be ‘joined up’ and compared.
The OECD DAC Creditor Reporting System (CRS) enables donor countries to account for their aid spending. It is divided into 26 sectors and 151 sub-sectors, which reflect donor countries’ priorities. The success of any intervention relies on the sustainability of the help as well as new solutions to old problems. CRS offers a sub-sector ‘malaria control’ under the umbrella of basic health, which entails prevention and control of malaria.
COFOG coding on the other hand deals with government spending data by the purpose for which the funds are used. The CRS sector ‘malaria’ does not have a counterpart in COFOG, although elements entailed can be broadly assigned to COFOG sub-sectors within the sector ‘health’ such as: ‘hospital services’, ‘medical products, appliances and equipment’, ‘outpatient services’, ‘public health services’ or ‘research and development in health’. Having a sector related to malaria in COFOG would have made it possible to assess whether funding levels could sustain the cost of treating malaria patients and ensure sufficient amounts of medication and preventative materials, such as mosquito nets, are available.
Conversely, if general OECD DAC sub-sector ‘malaria control’ could be broken down to the COFOG sub-sectors as listed above, a more detailed view of neglected areas could be discovered. Joining up data standards such as these can give us another layer of holistic understanding of the fate of international funds and help aid funds reach those most in need. USAID Forward has pledged for 30% of development assistance to go directly to local governments. But such a pledge can only be achieved in a transparent and efficient way when the standards are interoperable and joined up.
Linking data does provides data geeks such as me not only with an exciting challenge of ordering how the data is classified but also with an invaluable insight into the gaps in our understanding on how aid is distributed, received and where the gaps in our knowledge are. As the project continues and our online thesaurus grows, I will be raising more such questions.