Insights from the Data for Development Profiles of DAC members
Data and statistics provide the essential basis for understanding the practicalities of the development process, the interactions and feedbacks between different systems, and the factors that should shape decisions. They are vital for answering larger questions about the development process and identifying the reasons behind differential rates of growth, development and well-being. The supply of relevant, timely and usable data is essential for countries to set priorities, make informed choices and implement better policies for sustainable development. They are also a prerequisite for delivering on the 2030 Agenda for Sustainable Development and ensuring that no one is left behind (OECD, 2017[1]).
Yet, the quality, availability and timeliness of basic socio-economic and demographic data remain deficient in many parts of the developing world (Lange, 2020[2]). There is a large and increasing divide in data and statistical capacity for public policy across countries and this has only been underscored during the COVID-19 pandemic. On the one hand, data have become omnipresent in peoples’ lives in most advanced economies as they are used to decide, for instance, whether to re-open businesses or implement more stringent lockdowns. Key indicators of this pandemic – and with them the value of data for public policy – have become part of our collective consciousness. In many developing countries, on the other hand, the pandemic has served as a reminder of the lack of critical data, including basic data such as counts of the number of victims of the pandemic (BBC, 2021[3]) and data on its economic impact, resulting in a poor understanding of the impact of COVID-19 on people’s lives and the effectiveness of policy measures, especially in low-income and fragile states (Twiwwe and Wilkinson, 2020[4]).
Most members of the OECD’s Development Assistance Committee (DAC), along with multilateral organisations, private foundations and other stakeholders, actively engage in strengthening data and statistical systems in developing countries (Lange, 2020[2]). Yet building capacity to produce and use data and statistics effectively continues to pose significant challenges for providers of development co-operation. First, while providers of development co-operation often have an immediate need for data to inform their programming, monitor results and evaluate impact, robust and reliable data and statistical systems can take many years to establish. This can result in a trade-off between long-term investments that address countries’ needs and short-term, one-off data collection exercises that meet providers’ needs but build little capacity along the way. Second, an increasing number of partners, often with different priorities and mandates, renders effective co-ordination more challenging, especially in countries in which national statistical offices lack the capacity to co-ordinate support effectively (Lange, 2020[2]).
Third, new technologies, including digitisation and mobile connectivity, are creating new ways to source, disseminate and analyse data (PARIS21, 2020[5]) and uptake of these new technologies has increased significantly during the pandemic.1 But digital data also spark concerns over security, privacy and trust, especially in relation to data produced as a by-product of the use of digital services (World Bank, 2021[6]).2
Finally, official data and statistics are inherently political. When citizens have access to data and statistics and the literacy skills to use them, they play a central role in public accountability of governments and politicians, creating incentives for policy makers to manipulate data or keep them from being publicly available (Aragão and Linsi, 2020[7]; Dargent et al., 2018[8]). For providers of development co-operation, taking into account the political economy of data in a given country is important to ensure the effectiveness of support to data and statistics (Hoogeveen and Nguyen, 2019[9]) but this requires a solid understanding of political, cultural and historical contexts.
The Data for Development Profiles provide unprecedented detailed facts, figures and insights about individual DAC members’ data- and statistics-related ODA. They share information provided by DAC members on their motivations, objectives, strategies and lessons learnt as well as an analysis of the OECD’s aid flow data (see Box 1 and the methodological appendix for details on the data used for the profiles and in this chapter).
Official development assistance (ODA) for data and statistics: The data used in the profiles were extracted from the OECD’s aid flow database (OECD, 2020[10]) using three different approaches. First, all activities that used the designated purpose code for statistical capacity building were included. However, ODA to data production and use, including activities that support core functions of national statistical systems such as censuses and surveys, are not necessarily recorded under this purpose code and are recorded under a wide range of different purpose codes (PARIS21, 2019[11]). Other activities were identified based on a keyword search of project titles. The results were subsequently checked manually by the OECD and made available to Development Assistance Committee (DAC) members for fact-checking. Finally, in select cases and following guidance from the relevant members, members contributed additional data to fill gaps.
Unless otherwise noted, ODA for data and statistics refers to DAC members’ bilateral ODA, including earmarked funding channelled through multilateral organisations (also referred to as multi-bi aid). The ODA data do not cover DAC members’ core funding of multilateral organisations, which, by definition, cannot be matched to multilaterals’ development co-operation activities in support of data and statistics.
Data and statistical domains: ODA for data and statistics was classified as being in support of specific data and statistical domains modelled loosely on the Classification of Statistical Activities (UNECE, 2009[12]). Three criteria were used: 1) the aid purpose code under which activities were reported in the aid flow data; 2) select keywords; and, 3) in cases in which programmes and projects were implemented by multilateral organisations, the mandate of the implementing partner (e.g. data and statistical projects implemented by the International Monetary Fund [IMF] were assumed to be in support of economic data).
Limitations: It should be noted that there is no agreed-upon method to track ODA for data and statistics and there are limitations to the methodology employed. Importantly, as the OECD’s aid flow data do not provide details on programmes and projects beyond purpose codes, project titles and descriptions, reported activities that support data and statistics as a sub-component may or may not have been included in the analysis, depending on whether the project title had one of the keywords used. In some cases, DAC members highlighted key projects that the word search method missed and these projects were added. DAC members consulted for this research have indicated that more programmes and projects have data- and statistics-related components which may be missed in the analysis due to differences in reporting and levels of detail provided. It is probable that this analysis undercounts financial support for data and statistical production, capacity development, and other support to capabilities like data literacy, visualisation, registrations, etc.
The methodological appendix provides further details, including on partner-country characteristics (income groups and fragility status) and policy markers used to identify the objectives of ODA.
The countries profiled in this report cover more than 90% of DAC members’ total ODA to data and statistics from 2010 to 2019,3 and provide a solid basis for comparative analysis, identifying insights and key trends. This chapter highlights five insights on the priorities, strategies, and trends in allocations by partner country, theme and sector, as well as programming modalities. They also share lessons from evaluations and practical experience (see Infographic). The chapter also explores some of the implications for the future of international support for data and statistical systems of the insights and trends identified across profiles.
Insight # 1: DAC members’ overarching thematic and geographic priorities drive funding of data and statistics
With the exception of Italy, Portugal and Sweden, overarching development co-operation strategies do not identify strengthening data and statistical capacity as a strategic priority. Sweden’s 2016 Policy Framework for Swedish Development Cooperation and Humanitarian Assistance (Government of Sweden, 2016[13]), for instance, defines helping to improve countries’ own statistics systems as an objective in order to increase openness and transparency in relation to the 2030 Agenda.
However, the profiles show that there is often a clear link between DAC members’ overarching policy priorities and the type of support for data and statistics that they provide (Figure 1). Australia, for instance, combines a clear regional focus with support to health and disability data in line with its overall development co-operation priorities. Canada focuses on population and health data to strengthen maternal and child health as part of its feminist development co-operation strategy. Sweden, a leader on advancing gender equality, provides training on gender statistics and supports UN Women’s programming on gender data. Switzerland’s commitment to enhancing information systems and producing disaggregated data is in line with its strategic focus on those at risk of being left behind.
Notes: Based on gross disbursements. Not all official development assistance flows reported to the OECD are screened and assigned policy markers. Refer to the individual country profiles for more information.
Source: Authors’ calculations based on OECD (2020[10]), Creditor Reporting System (CRS) (database), https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
Data-related support is spread across institutions
The data and statistical components of DAC members’ programmes and projects tend to be managed by different units and divisions within one organisation and, in some cases, by different national agencies.4 While DAC members often have a focal point for narrowly defined statistical capacity building, it is rare to have a dedicated team or focal point for all data and statistics-related programmes and projects supported by the country or development agency. Only the United Kingdom’s FCDO stands out for having a dedicated Development Data Team that co-ordinates activities with country offices and other government entities. The process of compiling the profiles also showed that issues related to the production, analysis and use of data are increasingly more relevant in projects that do not primarily aim to strengthen data and statistical capacity (see also Insight # 2).
Digital strategies highlight data challenges and opportunities
Nevertheless, DAC members’ approaches to strengthening data and statistical systems and to using data in programme and project management appear to become more strategic – especially in the light of digitalisation. For example:
In 2018, Norway launched Digitalisation for Development (Norwegian Ministry of Foreign Affairs, 2018[14]), a strategy for digitalisation for Norwegian development policy. Among other things, the strategy highlights the importance of strengthening public registers.
In 2017, the United Kingdom’s Foreign, Commonwealth & Development Office (FCDO) (previously the Department for International Development) commissioned a Decision-making and Data Use Landscaping study (DFID, 2018[15]) to take stock of how FCDO teams manage, access, analyse and use internal and external data and to further inform their strategic approach to data.
Data-driven approaches, for planning and programming, results monitoring and evaluation, are also deeply rooted in United States’ government agencies that manage development co-operation, including the United States Agency for International Development (USAID) and the Millennium Challenge Corporation. USAID’s Digital Strategy, for instance, is an agency-wide strategy that includes initiatives aimed at advancing partner countries’ capacity to create and use data in development and humanitarian assistance (USAID, 2020[16]) and its Considerations for Using Data Responsibly (USAID, 2019[17]) provides a framework for identifying and understanding risks associated with collecting, sharing and using data in USAID programmes.
Long-standing geographic priorities bear out in support for data and statistics
In addition to the apparent alignment between DAC members’ ODA to data and statistics and their overarching thematic priorities, DAC members’ long-standing regional priorities also bear out in their support for data and statistics. Figure 2 illustrates geographic priorities, as captured by DAC members’ country-allocable ODA for data and statistics, based on three examples from the profiles:
Australia’s country-allocable ODA is concentrated in Asia, particularly Southeast Asia and the Pacific. It has no significant country-level engagement in Africa, the Americas or Europe.
The European Union (EU), on the other hand, is the key provider of co-operation in the Western Balkans and the Eastern and Southern Mediterranean, where it engaged in the context of the EU’s Pre-accession Assistance programmes and the European Neighbourhood Policy. In the context of its development policy, the EU also increasingly invests in statistical systems in partner countries in sub-Saharan Africa.
The United Kingdom’s partner countries are concentrated in Eastern Africa and South Asia, although it also has a few partner countries in Southern and Western Africa.
Note: Based on gross disbursements. Bubble sizes are proportional to members’ gross disbursements.
Source: Authors’ calculations based on OECD (2020[10]), Creditor Reporting System (CRS) (database), https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
Insight # 2: DAC members increasingly invest in sectoral data and statistics
Based on the data compiled for the profiles, DAC members’ combined ODA to data and statistics – capacity building and other forms of support for data and statistics – increased in real terms, albeit only moderately, from USD 276 million in 2010 to USD 339 million in 2019 – an annual rate of growth of 2.3%. The trend for DAC members is broadly in line with findings from the most recent Partner Report on Support to Statistics (PRESS) by PARIS21, which estimates a total of USD 672 million in 2019 for all providers of external funding (including multilateral organisations and private foundations) and no increase in total ODA over the last ten years (PARIS21, 2020[18]).
Figure 3 shows that the moderate increase in funding for data and statistics by DAC members was not driven by increased investment in core statistical capacity, but by funding of data and statistics in other domains. While it still accounted for the largest share of ODA for data and statistics in 2019, support to general statistical capacity building declined by more than 20%, from USD 105 million in 2010 (in 2018 prices) to about USD 82 million by 2019.
Note: Based on gross disbursements.
Source: Authors’ calculations based on OECD (2020[10]) Creditor Reporting System (CRS) Aid Activity Database, https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
A closer look at evolutions in sectoral support
Disbursements to population data and statistics nearly doubled between 2010 and 2018, from USD 40 million to USD 76 million in 2018, before declining again to USD 51 million in 2019. ODA to population data tends to be cyclical, with peaks towards the end of a decade when many countries field population censuses and troughs mid-decade. At the same time, the increase in funding for population data after 2015 is also driven by increasing investment in civil registration and vital statistics systems (CRVS), which feature prominently in the SDGs (e.g. SDG Target 16.9: By 2030, provide legal identity for all, including birth registration). Canada, Italy and Switzerland explicitly note CRVS as a priority in their profile. Starting in 2014, Canada, for instance, has scaled up substantially its funding of activities in support of maternal, newborn and child health as well as sexual and reproductive health rights – which often entails investments in CRVS systems.
An increase in funding of population data is also to some extent driven by a greater focus on issues around security and migration. To give one example, Denmark’s development co-operation policy positions as priorities the respective intersections between development and security and development and migration. In line with this, Denmark has increased its investment in migration data, captured in the profiles as a subset of population data, by supporting the World Bank-UNHCR Joint Data Center on Forced Displacement.
ODA for health data and statistics increased significantly between 2010 and 2015, from USD 35 million in 2010 to USD 73 million in 2015, but has since levelled off. The increase was driven mainly by the United States Agency for International Development (USAID), which increased investments in health data in partner countries by about USD 20 million over the entire period. Australia and the United Kingdom have also increased their investment in health data and statistics: Australia expanded its investment in health data in partnership with Bloomberg Philanthropies, while the United Kingdom frequently supported health and nutrition surveys in its partner countries.
Funding for gender and environmental statistics also increased after 2015, by USD 6 million and USD 5.7 million per year, respectively (see Figure 3). However, in both cases, the surge in funding started from a much lower base – close to zero in the case of gender statistics. Gender data and statistics are a key priority for Australia, Canada and Sweden, which, along with Ireland (not profiled), the United Kingdom, the United States and other partners support UN Women’s Women Count programme. Sweden in 2016 also launched an International Training Programme in Gender Statistics to support partner countries’ capacity to produce and use gender statistics.
ODA to strengthen economic statistics doubled between 2010 and 2019, from around USD 20 million to USD 40 million by 2019. The increase is, to a significant extent, driven by co-operation between DAC members and the IMF, for instance, in the context of the IMF’s Data for Decisions (D4D) Fund.
While the profiles do not attempt to quantify ODA invested in new data such as “Big Data” or remote-sensing data, a survey of DAC members in 2017 found that nine DAC members had already started looking into the potential contribution of Big Data to development co-operation while six members were thinking about working with Big Data (Sanna and Mc Donnell, 2017[19]). Several profiles indicate that DAC members are investing in innovative data:
In 2019, Australia helped establish Digital Earth (DE) Africa, a Geoscience Australia digital platform for the use of satellite information to address sustainable development challenges.
Australia, Denmark and Sweden have all provided support in the past to the United Nations Global Pulse, the UN Secretary-General’s initiative on Big Data and artificial intelligence for development, humanitarian action and peace.
The United Kingdom’s FCDO has teamed up with the Met Office, the National Aeronautics and Space Administration (NASA) in the United States and US scientists to use NASA satellite data to accurately predict where and when cholera will spread. The United Kingdom also supports governments to collect, use and share geospatial data on population settlement and infrastructure through its Geo-referenced Infrastructure and Demographic Data for Development (GRID3) programme.
Programmatic and budget support can incentivise support for general/core statistical capacities
While changes in policy objectives and priorities during the SDG era have an impact on data demand, DAC members also note the role of aid modalities – budget support, in particular – in creating incentives to support certain types of data and statistics. Budget support, an aid modality that relies on direct funding to the recipient government’s treasury, combined with high-level policy dialogue and shared monitoring and results frameworks, seems to create demand for key development indicators that can be produced by the national statistical system (Box 2). Two profiles suggest that this demand also creates incentives to invest in general statistical capacity building:
The United Kingdom notes that budget support, which it relied on extensively during the first decade of the century, strengthened country offices’ rationale for investing in statistical capacity development in partner countries to help them monitor progress towards national objectives. The focus shifted towards centrally managed programmes and sectoral programmes supported by country offices after the United Kingdom’s withdrawal from budget support starting in 2010.
The European Union, which still makes use of budget support as a modality, notes in its profile that its statistical capacity support often focuses on key economic and societal variables as these are often needed as performance indicators in budget support programmes.
While the evidence on the link between budget support and incentives to invest in core statistical systems remains anecdotal, the shift away from funding of general statistical capacity building and towards sectoral support for data and statistics is consistent with the withdrawal of European DAC members from budget support over the course of the 2010s.
The Global Partnership for Effective Development Co-operation tracks progress against Sustainable Development Goal (SDG) Target 17.15 on respecting countries’ policy space and leadership to establish and implement policies for the SDGs. This includes tracking uptake among providers of development co-operation of government data and statistics for results monitoring. Yet, while Development Assistance Committee (DAC) members in 2018 on average aligned 80% of their project objectives to partner country priorities, they used national data and statistics for their result monitoring only 41% of the time, down from 50% in 2016 (GPEDC, 2019[20]).
Further analysis of data from the 2016 and 2018 Global Partnership monitoring rounds shows that the share of DAC members’ disbursements to the public sector is a key determinant of use of national data and statistics for project monitoring: a ten percentage point increase in the share of disbursements “on budget” is associated with a four percentage point increase in the share of results indicators drawn from government sources and monitoring systems (Figure 4). This is in line with the notion that direct funding of partner governments creates demand for government data.
Note: Solid line based on an ordinary least squares regression. The shaded area indicates the 95%-confidence interval.
Source: Authors’ calculation based on GPEDC (2018[21]) data, https://www.effectivecooperation.org/landing-page/monitoring-data.
Insight # 3: ODA to data and statistics is increasingly concentrated in fragile states
ODA is increasingly concentrated in Africa...
Despite the long-standing regional priorities noted earlier (Insight # 1), DAC members annual ODA to data and statistics allocated directly to partners in Africa increased by nearly USD 44 million (in 2018 prices) between 2010 and 2019, from USD 88 million to USD 132 million (Figure 5). In 2017-19, Africa accounted for about 40% of DAC members’ total ODA to data and statistics, up from 32% at the beginning of the decade. Over the same period, annual bilateral ODA for data and statistics to partners in the Americas increased only modestly (from USD 18 million per year in 2010 to USD 31 million in 2019). It remained flat in Asia and Oceania at about USD 65 and USD 5 million per year, and decreased in Europe, from USD 38 million to USD 22 million annually. Finally, there is a marked uptick in funding channelled through global programmes and initiatives in the SDG era, with funding increasing from USD 46 million in 2014 to USD 98 million by 2019.
Note: Based on gross disbursements allocable at the regional level.
Source: Authors’ calculations based on OECD (2020[10]), Creditor Reporting System (CRS) (database), https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
Given that many countries in sub-Saharan Africa continue to lag in the availability of key data (BBC, 2021[3]; Devarajan, 2013[22]; Hoogeveen and Nguyen, 2019[9]), the increasing focus on the region would appear to be concentrating resources where need is the greatest. However, as an increasing number of providers of development co-operation support the same set of partner countries, it also potentially increases the need for better co-ordination among these countries while raising the question of whether some regions or countries are being left behind.
...and in fragile contexts
Out of the two-thirds of DAC members’ ODA directed to specific countries,close to 80% is targeted to low-income (LICs) and lower middle-income countries (LMICs) (as defined in 2019). The share disbursed to today’s LICs increased, from 37% in 2015 to 50% in 2016, before falling again to 42% in 2019 (Figure 6). Country-specific bilateral ODA to data and statistics is also increasingly concentrated in countries and territories classified as fragile: in 2019, 69% was targeted to countries classified as fragile, up from 46% in 2011. DAC members’ ODA to data and statistics is also increasingly concentrated in countries with weaker governance and in which citizens have fewer ways of expressing their views and holding government to account, with potential implications for the effectiveness of their support (Box 3).
Note: Based on gross disbursements allocable to individual countries. Country income groups and fragility status are fixed over time and are those applied to reporting to the OECD in 2019. Left panel: LIC: low-income country; LMIC: lower middle-income country; UMIC: upper middle-income country; HIC: high-income country.
Source: Authors’ calculations based on OECD (2020[10]), Creditor Reporting System (CRS) (database), https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
The shift towards greater investments in LICs and fragile contexts is especially apparent in the profiles of Canada, Japan and Korea, but also in those of the European Union, Norway and Sweden. This shift has often coincided with a reorientation of funding away from other regions towards Africa. Other DAC members such as the United Kingdom and the United States maintained a high share of funding to LICs and fragile contexts over the course of the 2010s.
Countries’ statistical capacity closely correlates with indicators of good governance – and development Assistance Committee (DAC) members’ bilateral official development assistance (ODA) to data and statistics is increasingly concentrated in countries with weaker governance and in which citizens have fewer ways of expressing their views and holding government to account (Figure 7). Based on the Worldwide Governance Indicators (Kaufmann and Kraay, 2020[23]) and data on DAC members’ country-allocable ODA to data and statistics, countries ranking among the bottom 40% in terms of government effectiveness in 2019 received more than 70% of ODA to data and statistics, up from 60% in 2010. And while the bottom 40% of countries in terms of voice and accountability received 51% of DAC members’ country-allocable ODA in 2010, by 2019, their share had also increased to 74%.
Note: Left panel based on gross disbursements. Z-scores are normalised values of two variables; log GNI per capita and Kaufmann and Kray’s indicator of government effectiveness. GNI: gross national income.
Source: Authors’ calculations based on OECD (2020[10]), Creditor Reporting System (CRS) Aid Activity Database, https://stats.oecd.org/Index.aspx?DataSetCode=CRS1, Kaufmann and Kray (2020[23]), Worldwide Governance Indicators, www.govindicators.org, and World Bank (2020[24]), Statistical Capacity Indicator, https://datatopics.worldbank.org/statisticalcapacity/.
Across countries, the last ten years have seen a decline in democracy (Alizada et al., 2021[25]), a trend that continued in 2020 and has brought with it a concentration of ODA in countries that face more severe governance challenges. The implications of both regime type and governance for aid effectiveness have long been debated (Burnside and Dollar, 2000[26]). But because of data’s potential to strengthen government effectiveness and accountability, there may be separate considerations in the area of support to data and statistics.
On the one hand, international support may help strengthen the accountability function of data and statistics. An example of this is the Eurasia Foundation’s TAPAS (Transparency and Accountability in Public Administration and Services) programme in Ukraine (Eurasia Foundation, n.d.[27]), funded by both the United Kingdom and the United States (see the United States’ profile). The programme aims to improve the efficiency and effectiveness of government and reduce corruption by making data on government contracts public on line, thus allowing civil society actors to monitor government procurement and raise questions if need be.
On the other hand, ensuring that support to government data and statistical systems is effective – accurate data and statistics are produced reliably, openly available and used by civil society – may be more challenging in these contexts. Countries in which democratic institutions are perceived to be under threat have in recent years made headlines for alleged political interference in the realm of official data and statistics (The Economist, 2019[28]; Wilson and Blood, 2019[29]) – although the phenomenon is by no means limited to autocracies (Aragão and Linsi, 2020[7]). Yet, if relatively small elites exert significant political, economic and military power, incentives to suppress or manipulate accurate official data and statistics grow stronger while the appearance of compliance with international standards for official statistics remains (Hoogeveen and Nguyen, 2019[9]). As traditional approaches to statistical capacity building are unlikely to address structural barriers to improving governance, different approaches to supporting data and statistics may be needed.
Insight # 4: Project-type interventions are preferred to joint-funding mechanisms
The share of DAC members’ ODA to data and statistics delivered in the form of joint-funding mechanisms – core support, contributions to specific programmes, basket funds, etc. – decreased from around 30% at the beginning of the 2010s to a little more than 20% in the second half of the decade (Figure 8). At the same time, project-type interventions, i.e. interventions associated with specific inputs, activities and outputs, increased their share from 54% in 2010 to 69% in 2019. Project-type interventions account for more than 50% of all ODA for 10 out of the 14 DAC members profiled in this report, including major sources of funding such as the European Union, Korea and the United States.
Note: Based on gross disbursements. NGO: non-governmental organisation; PPP: public-private partnership; TA: technical assistance.
Source: Authors’ calculations based on OECD (2020[10]) Creditor Reporting System (CRS) Aid Activity Database, https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
An increase in the volume of ODA to data and statistics delivered in the form of project-type interventions, from USD 150 million in 2010 to USD 237 million in 2019 (in 2018 prices), explains much of the increase in their share. The European Union (+USD 28.4 million per year), Korea (+11.1), the United Kingdom (+12.6), and the United States (+32.6) together account for nearly the entire increase of USD 87 million per year. However, this is not a universal trend: Japan, Sweden and Switzerland have reduced the share of their ODA disbursements delivered in the form of project-type interventions.
Note: Based on gross disbursements.
Source: Authors’ calculations based on OECD (2020[10]), Creditor Reporting System (CRS) Aid Activity Database, https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
Stable but evolving engagement in joint funding mechanisms
Funding in the form of contributions to specific programmes and funds of implementing partners have overall remained nearly constant between 2010 and 2019 at around USD 63 million per year. However, use of this modality has still changed over the decade. It was dominated by the United Kingdom’s contributions to trust funds and specific programmes, notably the World Bank’s Statistics for Results Facility and Catalytic Fund and the Trust Fund for Statistical Capacity Building, two trust funds dedicated to strengthening statistical capacity development.5 Towards the late 2010s, this modality was used more prominently by a larger number of DAC members and for different purposes. Examples include UN Women’s flagship programme Making Every Woman and Girl Count (from 2016, focus on gender data and statistics), the International Monetary Fund’s Data for Decisions Fund (from 2018, economic statistics) and Denmark’s support for the World Bank-UNHCR Joint Data Center on Forced Displacement (from 2018, migration data and statistics).
Basket funds,6 a funding mechanism by which providers of development co-operation contribute financial resources to an autonomous account managed jointly with other providers and/or the recipient, have played a minor role throughout the 2010-19 period. However, they have been important in select partner countries. Examples include the Common Fund of the Mozambique National Institute of Statistics (supported by Canada, Denmark, Italy, Norway, Portugal, Sweden and the United Kingdom) and funding arranged to support the modernisation of Rwanda’s National Institute of Statistics and its 2009/10-13/14 national strategy for the development of statistics (European Union and United Kingdom). DAC members’ ODA disbursements under this modality peaked in 2012. The Common Fund of Mozambique’s National Institute of Statistics, for instance, phased out after 2017 (Sida, 2019[30]).
Finally, a large share of DAC members provide funding for experts (often from their own statistical offices) and technical assistance. Canada, the European Union, Norway, the United Kingdom and the United States jointly accounted for two thirds to three fourths of funding in this category over the 2010-19 period. Among DAC members’ profiled, Portugal, which delivers nearly 90% of its support in the form of experts and technical assistance, tops the list in terms of relative use of this modality.7
Multilateral channels are often used in low-income countries and fragile contexts
DAC members rely on four key delivery channels to provide their financial assistance to data and statistics in partner countries (Figure 10): 1) funding channelled through the multilateral system (i.e. multi-bi); 2) funding channelled through DAC member public sector entities (often their national statistical offices); 3) direct funding of partner country governments; and 4) funding of interventions by private sector entities. Their respective shares have remained constant over the last decade. Multi-bi funding has accounted for one-third of DAC members’ total ODA for data and statistics since 2010, while funding of public sector entities has accounted for another third, split roughly between DAC member public sector entities and recipient governments. The United States channels a large fraction of its funding through the private sector. Research and teaching institutions and non-governmental organisations each account for 5% of total DAC ODA for data and statistics.
DAC members are more likely to provide earmarked funding via multilateral channels in LICs and fragile states while funding and technical assistance channelled through DAC members’ public sector entities and directly to recipient governments is less common in these countries (Figure 10). In general, funds are earmarked for various reasons, such as increasing the visibility of countries’ contributions, improving accountability on the use of funds towards taxpayers, fulfilling pledges to support specific causes or ensuring support to priorities that are deemed underfunded (OECD, 2020[31]; Bosch, Fabregas and Fisher, 2020[32]). Half of all country-specific activities in LICs implemented by multilaterals are project-type interventions rather than programmatic earmarking, which would allow for greater flexibility for implementing partners to align with their priorities. The share of project-type interventions is larger still in fragile contexts.
Notes: Based on gross disbursements. DAC: Development Assistance Committee; NGO: non-governmental organisation; LIC: low-income country; LMIC: lower middle-income country; UMIC: upper middle-income country. The distinction between programmatic and project-type earmarked funding to multilaterals follows Bosch, Fabregas and Fisher (2020[32]).
Source: Authors’ calculations based on OECD (2020[10]), Creditor Reporting System (CRS) (database), https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
Insight # 5: Country ownership drives successful results
The development of sustainable capacity, including data and statistical capacity, is a major challenge (OECD, 2008[33]) that requires the willingness to learn from experience and make appropriate adjustments along the way. Dedicated or stand-alone capacity building support to national statistical systems is a relatively recent area for development co-operation, with major initiatives taking off only in the late 1990s8 (Lange, 2020[2]) and a body of evaluations of different initiatives building up over the course of the 2000s (Willoughby, 2008[34]). DAC members have since accumulated a wealth of experience and lessons on what works in statistical capacity building. The profiles show which countries have expertise with specific engagement models, identify lessons and experience for peer learning, and validate their own experience. In particular, the profiles identify good practices and insights on lessons learnt and how support is evolving.
Regional programmes are emerging as relevant and efficient
DAC members highlight the advantages of specific engagement models, such as long-term technical assistance and support of regional programmes. For example, DAC members whose engagement model builds on partnerships between development co-operation agencies and their own national statistical offices note the advantages of providing long-term, flexible technical assistance, especially in fragile contexts, to ensure sustained capacity development. Members that support regional programmes, notably Australia, Canada and the European Union, note that programmes for countries in similar situations make it easier to ensure their relevance, foster peer learning and can turn into a stepping-stone towards South-South co-operation.
Approaches to supporting country ownership are evolving, going beyond national government
Country ownership, the notion that partner countries should exercise effective leadership over their development strategies and policies and take a lead in co-ordinating development actions, is a key principle of the Paris Declaration on Aid Effectiveness (OECD, 2005[35]). Country ownership is also seen as crucial by DAC members in the context of support to data and statistical capacity in order to increase the probability of sustainable results beyond project timelines. Lessons shared by key providers of technical assistance, in particular the European Union, Norway, Portugal and Sweden, all note that ensuring that support is in line with partner countries’ own priorities and strategies is crucial to achieving long-lasting development of partners’ statistical systems. The United States’ Millennium Challenge Corporation notes the potential ineffectiveness of parallel data collection for programme design, as it may undermine the uptake and development of local data systems.
Enabling greater ownership entails tailoring interventions to partners’ needs to ensure that there will be local demand and use for the data and statistics, especially for policy making. Greater ownership could be enabled by strengthening existing data systems, including at subnational levels, rather than collecting programme-specific data for ODA-funded projects and programmes from scratch. Yet these may not always be equipped to produce the data that providers of development co-operation seek (see Box 4). Insufficient core funding of NSOs and other data-producing government entities by their governments poses additional challenges, as it can curtail capacity to absorb international support for data and statistics while also providing incentives to agree to whatever support is offered (Lange, 2020[2]). A sound understanding of the context in which support to data and statistics is provided – the institutions as well as the structures of power and influence – is key to designing programmes and projects that partners’ ‘own’.
Political economy considerations (see Box 3) aside, governments need data that allow them to respond effectively to the needs of their citizens. Is there a need to hire more teachers? Or should there be more investment in school infrastructure? Importantly, for governments to be able to act upon statistical information, they need data that can be disaggregated at the relevant administrative level (e.g. school districts) (Rosling and Fleck, 2013[36]). These considerations will often point to administrative data collected in the process of providing government services.
Providers of development co-operation, on the other hand, will often be interested in international comparisons, for instance of progress towards the international development goals (OPM, 2009[37]; Sandefur, 2013[38]). Standardised household surveys, which provide a high degree of comparability across countries, are often well-suited for this purpose. At the same time, there has been an increasing focus on the accountability and reporting function of results-based management in development co-operation (Zwart and Egan, 2017[39]; Vähämäki and Verger, 2019[40]), which created additional demand for data for monitoring results. Interventions that are targeted to very specific sections of the population, for instance, might require data that are not readily available from government sources, especially in countries with low levels of statistical capacity, or data that are collected at very specific moments in time (e.g. before and after an intervention takes place). This points to specialised surveys, with limited incentives to collect data on issues that are not immediately relevant to the project.
Governments will be interested in ensuring that surveys are relevant to country context and less so in their usefulness in international comparisons. In fact, governments and official statisticians often highlight the importance of administrative data (PARIS21, UNSD, 2018[41]), which are less likely to be useful for international comparisons. Small-scale surveys for project monitoring and evaluation, on the other hand, will be useful only to the stakeholders of the project, even if the resulting data are made available to other stakeholders. These differences in information needs can lead to activities in support for data and statistics that are not well-aligned with long-term country priorities – that are not “owned” by developing country governments.
The country profiles serve as a baseline for strategic international dialogue for more co-ordinated and effective support that contributes to strengthened capacity and statistical systems fit for the digital era. They can also inform DAC members future data strategies, flagging the need for more holistic and whole-of-government investments; the benefits to partners of following good practices that are informed by the peer learning and experience so far with increasing effectiveness for more sustainable results; and pursuing more harmonised programmes and capacity development. Three particular insights emerge from the profiles which can shape future strategies:
1. Increase understanding and manage the tensions and potential trade-offs between different drivers of ODA investments in data and statistics for sustainable development and for effective development co-operation. Data and statistics are a means to ensure transparency and mutual accountability of development co-operation (Zwart and Egan, 2017[39]). Providers of development co-operation play a dual role in relation to partner countries’ data ecosystems: they invest in building and strengthening national statistical systems and as users of development data they invest in data production for bilateral projects and programmes, results monitoring and evaluation; often in parallel to national systems. Tensions, or inconsistencies, arise between donor pressure to gather data to show results of on-going programmes and projects and the commitments to rely on country data and use and build up country systems. In their role as denizens of countries’ data ecosystems, development co-operation providers should ensure that their demand for data supports the development of local capacities or, at a minimum, that their ambition to be data-driven does not undermine local systems. Strategies for development co-operation for data and statistics need to reflect a better understanding and awareness of the potential trade-offs between the different data roles providers play and provide guidance on how to resolve tensions effectively.
2. Prioritise inter and intra-institutional co-ordination and coherence domestically to find more synergies and more coherent agency-wide support for national statistical systems. Providers should take a principled, coherent approach to supporting data and statistical systems. At present, few DAC members have a dedicated team or focal point mandated to monitor and provide guidance on all data and statistics-related programmes and projects. Comprehensive strategies on how best to support data and statistical systems are also lacking. To identify and leverage synergies, agency-wide strategies for data and statistics could be developed that raise awareness of the importance of sound data and statistical systems in partner countries, reflect adequately the public-good nature of data, and provide guidance on how best to support partners’ data systems in a holistic manner.
3. Identify and adopt international good practices that result in more coherent and effective international co-operation, that draw on peer learning and lessons, and that promote more aligned and harmonised co-operation for data for development. To ensure coherence in development co-operation, strategies for data and statistics should aim to support co-ordination among providers at the design, planning and the implementation stages. The Data for Development Profiles indicate some good practices such as the importance of a sound understanding of country contexts, upholding the principle of country ownership, and engaging in long-term, flexible support for data and statistical systems.
References
[25] Alizada, N. et al. (2021), Democracy Report 2021: Autocratization turns viral, University of Gothenburg: V-Dem Institute, Gothenburg, https://www.v-dem.net/files/25/DR%202021.pdf (accessed on 7 April 2021).
[7] Aragão, R. and L. Linsi (2020), “Many shades of wrong: What governments do when they manipulate statistics”, Review of International Political Economy, https://doi.org/10.1080/09692290.2020.1769704.
[3] BBC (2021), “Measuring Africa’s data gap: The cost of not counting the dead”, BBC, https://www.bbc.com/news/world-africa-55674139 (accessed on 29 March 2021).
[32] Bosch, E., A. Fabregas and F. Fisher (2020), Earmarked Funding to Multilateral Organisations: How Is It Used and What Constitutes Good Practice?, OECD, Paris, https://www.oecd.org/dac/financing-sustainable-development/development-finance-topics/Multilateral-development-finance-brief-2020.pdf.
[26] Burnside, C. and D. Dollar (2000), “Aid, Policies, and Growth”, American Economic Review, Vol. 90/4, https://doi.org/10.1257/aer.90.4.847.
[8] Dargent, E. et al. (2018), Who Wants to Know? The Political Economy of Statistical Capacity in Latin America, Inter-American Development Bank, Washington, DC, https://publications.iadb.org/en/who-wants-know-political-economy-statistical-capacity-latin-america.
[22] Devarajan, S. (2013), “Africa’s statistical tragedy”, Review of Income and Wealth, Vol. 59/S1, pp. S9-S15, https://doi.org/10.1111/roiw.12013.
[15] DFID (2018), Decision-making and Data Use Landscaping: Better Data, Better Decisions, Department for International Development and Development Gateway, https://developmentgateway.org/wp-content/uploads/2020/10/Better_Data_Better_Decisions_Data_Landscape_Study-1.pdf.
[27] Eurasia Foundation (n.d.), Transparency and Accoutability in Public Administration and Services.
[13] Government of Sweden (2016), Policy Framework for Swedish Development Cooperation and Humanitarian Assistance, Government of Sweden, Stockholm, https://www.government.se/49a184/contentassets/43972c7f81c34d51a82e6a7502860895/skr-60-engelsk-version_web.pdf.
[20] GPEDC (2019), Making Development Co-operation More Effective - 2019 Progress Report, OECD, UNDP, https://www.oecd-ilibrary.org/development/making-development-co-operation-more-effective_26f2638f-en.
[21] GPEDC (2018), GPEDC Monitoring Data, https://www.effectivecooperation.org/landing-page/monitoring-data (accessed on 8 April 2021).
[9] Hoogeveen, J. and N. Nguyen (2019), “Statistics reform in Africa: Aligning incentives with results”, Journal of Development Studies, Vol. 55/4, pp. 102-179, https://doi.org/10.1080/00220388.2017.1417583.
[23] Kaufmann, D. and A. Kraay (2020), www.govindicators.org, Worldwide Goernance Indicators.
[2] Lange, S. (2020), “Key trends in development co-operation for national data and statistical systems”, OECD Development Policy Papers, No. 31, OECD Publishing, Paris, https://doi.org/10.1787/1ce044d2-en.
[14] Norwegian Ministry of Foreign Affairs (2018), Digitalisation for Development: Digital Strategy for Norwegian Development Policy, Norwegian Ministry of Foreign Affairs, Oslo, https://www.regjeringen.no/globalassets/departementene/ud/dokumenter/utvpolitikk/digital_strategynew.pdf (accessed on 18 March 2021).
[10] OECD (2020), Creditor Reporting System (CRS) (database), OECD, Paris, https://stats.oecd.org/Index.aspx?DataSetCode=CRS1.
[31] OECD (2020), Multilateral Development Finance 2020, OECD Publishing, Paris, https://dx.doi.org/10.1787/e61fdf00-en.
[1] OECD (2017), Development Co-operation Report 2017: Data for Development, OECD Publishing, Paris, https://dx.doi.org/10.1787/dcr-2017-en.
[33] OECD (2008), “The challenge of capacity development: Working towards good practice”, OECD Journal on Development, Vol. 8/3, OECD Publishing, Paris, https://dx.doi.org/10.1787/journal_dev-v8-art40-en.
[35] OECD (2005), Paris Declaration on Aid Effectiveness, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264098084-en.
[37] OPM (2009), Evaluation of the Implementation of the Paris Declaration: Thematic Study – Support to Statistical Capacity Building, Synthesis Report, http://www.oecd.org/dac/evaluation/dcdndep/43189996.pdf.
[5] PARIS21 (2020), Guidelines for Developing Statistical Capacity. A Roadmap for Capacity Development 4.0, PARIS21, Paris, https://paris21.org/sites/default/files/inline-files/UNV003_Guidelines%20for%20Capacity%20Development%20PRINT_0.pdf.
[18] PARIS21 (2020), The Partner Report on Support to Statistics 2020, Partnership in Statistics for Development in the 21st Century, Paris, https://paris21.org/sites/default/files/inline-files/PRESS%202020%20Final.pdf.
[11] PARIS21 (2019), “Proposing a New Approach to Measure Support to Statistics and Data in the OECD Creditor Reporting System”, OECD, Paris, France, http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=DCD/DAC/STAT(2019)26&docLanguage=En.
[41] PARIS21, UNSD (2018), Joint survey on new approaches to capacity develompent and future priorities, Survey of national statistical offices, https://paris21.org/capacity-development-40/cd40-survey (accessed on 12 April 2021).
[36] Rosling, H. and F. Fleck (2013), “The joy of facts and figures”, Bulletin of the World Health Organization, Vol. 91/12, https://doi.org/10.2471/BLT.13.031213.
[38] Sandefur, J. (2013), Seeing like a state in Africa: Data Needed, Center for Global Develpment blog post, https://www.cgdev.org/blog/seeing-state-africa-data-needed (accessed on 7 April 2021).
[19] Sanna, V. and I. Mc Donnell (2017), “Data for development: DAC member priorities and challenges”, OECD Development Co-operation Working Papers, No. 35, OECD Publishing, Paris, https://www.oecd.org/dac/WP35%20Complete.pdf.
[30] Sida (2019), Review of Scandinavian Consortium Technical Assistance to Strengthening the Institutional Capacity of the Mozambican National Statistical System (2008–2017), Swedish International Development Agency, Stockholm, https://publikationer.sida.se/English/publications/163496/review-of-scandinavian-consortium-technical-assistance-to-strengthening-the-institutional-capacity-of-the-mozambican-national-statistical-system-20082017.
[28] The Economist (2019), “A dose of bull”, The Economist, https://www.economist.com/middle-east-and-africa/2019/03/14/tanzanias-leader-the-bulldozer-runs-off-course (accessed on 7 April 2021).
[4] Twiwwe, S. and N. Wilkinson (2020), “Evidence for policy making in uncertain times”, in Development Co-operation Report 2020: Learning from Crises, Building Resilience, OECD Publishing, Paris.
[12] UNECE (2009), Classification of Statistical Activities, International standard, https://statswiki.unece.org/display/CSA/Classification+of+Statistical+Activities (accessed on 12 April 2021).
[42] UNFPA (2021), “Technical brief on the implications of COVID-19 on Census”, United Nations Population Fund, New York, https://www.unfpa.org/resources/technical-brief-implications-covid-19-census (accessed on 15 April 2021).
[16] USAID (2020), Digital Strategy 2020-2024, United States Agency for International Development, Washington, DC, https://www.usaid.gov/sites/default/files/documents/15396/USAID_Digital_Strategy.pdf.
[17] USAID (2019), Considerations for Using Data Responsibly at USAID, United States Agency for International Development, Washington, DC, https://www.usaid.gov/responsibledata (accessed on 12 April 2021).
[40] Vähämäki, J. and C. Verger (2019), “Learning from Results-based management evaluations and reviews”, OECD Development Co-operation Working Paper, No. 53, OECD, Paris.
[34] Willoughby, C. (2008), “Overview of evaluations of large-scale statistical capacity building initiatives”, The Partnership in Statistics for Development in the 21st Century, Paris, https://paris21.org/sites/default/files/3640.pdf (accessed on 13 April 2021).
[29] Wilson, T. and D. Blood (2019), “Rwanda: where even poverty data must toe Kagame’s line”, Financial Times, https://www.ft.com/content/683047ac-b857-11e9-96bd-8e884d3ea203 (accessed on 7 April 2021).
[6] World Bank (2021), World Development Report 2021: Data for Better Lives, World Bank, Washington, DC, https://www.worldbank.org/en/publication/wdr2021 (accessed on 19 June 2020).
[24] World Bank (2020), Data on Statistical Capacity, http://datatopics.worldbank.org/statisticalcapacity/.
[39] Zwart, R. and J. Egan (2017), “Making better use of results data in development co-operation”, in Development Co-operation Report 2017, OECD Publishing, Paris.
Notes
← 1. A general trend among national statistical authorities towards greater reliance on both administrative data and new data sources has accelerated as COVID-19 increased the demand for timely and granular data and led to the postponement or cancellation of traditional data collection exercises such as household surveys and censuses (UNFPA, 2021[42]).
← 2. The World Bank’s 2021 World Development: Report Data for Better Lives calls for a new “social contract for data”, starting with a renewed effort to improve data governance domestically and through closer international co-operation (World Bank, 2021[6]).
← 3. The 14 DAC members profiled account for about 94% of total ODA to data and statistics from DAC members in any given year.
← 4. Funding for economic statistics via the IMF, for instance, often originates outside of the main development co-operation agencies, such as in the case of Germany, Japan and Korea, where co-operation with the IMF falls under the purview of the respective Ministry of Finance, or in the case of Switzerland, where the State Secretariat of Economic Affairs funds IMF capacity development programmes.
← 5. These funds reached their closing dates in 2019 and 2020, respectively.
← 6. Basket funds are a funding mechanism by which providers of development co-operation contribute funds to an autonomous account, managed jointly with other donors and/or the recipient. The account will have specific purposes, modes of disbursement and accountability mechanisms, and a limited time frame. Basket funds are characterised by common project documents, common funding contracts and common reporting/audit procedures with all donors. There are some inconsistencies across reporters supporting the same activities. Luxembourg, for instance, reports contributions to the International Monetary Fund’s Data for Decisions Fund under this modality while other DAC members typically report it as a contribution to a specific-purpose fund/programme managed by an implementing partner. The modality is also often identified with pooled funding for the implementation of population and housing censuses.
← 7. As the distinction between „experts and technical assistance“ and „project-type interventions“ is not always clear, it is difficult to interpret the trends. The United States, for instance, notes that technical assistance is often a component within its project-type support. And the European Union’s MEDSTAT programme was classified under „experts and technical assistance” until 2015 when the fourth instalment of it would be classified as a „project-type intervention“.
← 8. The DAC, for instance, established a dedicated purpose code for statistical capacity building only in the mid-1990s.