2.1. Technical progress would enable greater data supply and use for SDG monitoring (Assumption 1)

The first assumption was that recent technological advances would exponentially increase the volume, velocity, veracity, and variety of data supply (Lucas, Reference Lucas2015; Melamed, Reference Melamed2014). By 2014, the global development community had begun using a wide range of new data collection tools such as mobile phone data and spatial population data for various crises, including the West African Ebola outbreak (USAID, 2017), as well as crowd-sourced infrastructure mapping for development and humanitarian platforms (Mooney and Minghini, Reference Mooney and Minghini2017). This was accompanied by an expectation of a surge in “big data,” from mobile phone data to earth observation data to social media data (Letouze, Reference Letouze2012), crowdsourcing, and citizen-science (Fraisl et al., Reference Fraisl, See, Campbell, Danielsen and Andrianandrasana2023; Grossman et al., Reference Grossman, Platas and Rodden2018). There was also an expectation that there would be more open data and more data sharing and reconciliation of administrative data (Rodriguez and Schonrock, Reference Rodriguez and Schonrock2018) and private sector data (GPSDD, 2018). Policies encouraging open statistics and data sharing further supported this trend (ODW, 2015).

Subsequent experience has reaffirmed this assumption. Data supply has increased, impacting SDG reporting since the WTC’s publication. Although aggregate data supply growth is not reliably measured, technology firms estimate a 5,000% increase in data creation and storage between 2010 and 2020, alongside growth in internet users, broadband, mobile connectivity, and administrative data for public services (Press, Reference Press2020). The velocity of data collection has also increased, leveraged in global SDG data initiatives like Data4Now, which enables countries to generate and utilize near real-time data for policy and monitoring (Espey, Reference Espey2019b). SDGs Today is another collaboration visualizing and reporting near real-time SDG indicators, some updated hourly (SDSN, 2023).

While a full systematic assessment is needed to uncover specific patterns of innovations for each SDG target, a rapid review since 2015 highlights multiple examples of new technologies being used to increase data collection. Monitoring poverty levels under SDG 1 has advanced using remote sensing and satellite imagery to supplement survey data for better estimates and predictive analysis of disaggregated poverty rates (Andree et al., Reference Andree, Mahler and Newhouse2023). Health and wellbeing under SDG 3 have benefited from mobile phone data to predict infectious disease spread and measure service coverage (Dahmm, Reference Dahmm2020; Oliver et al., Reference Oliver, Oliver, Lepri, Sterly, Lambiotte, Deletaille, De Nadai, Letouzé, Salah, Benjamins, Cattuto, Colizza, de Cordes, Fraiberger, Koebe, Lehmann, Murillo, Pentland, Pham and Vinck2020). Achieving zero hunger (SDG 2) involves measuring food prices using online and crowd-sourced data sources to assess food security in near-real time. Clean water and sanitation access (SDG 6) is being measured using water utility sensors, satellite imagery, and administrative data to improve global measures (Fischer, Reference Fischer2017; Thomson and Koehler, Reference Thomson and Koehler2016). Maritime vessel tracking data from automatic identification systems has revealed illegal fishing activities impacting SDG 14 (Coleman, Reference Coleman2022). These examples demonstrate significant advances in data supply for various SDG targets, with much of it leveraged by government services and civil society organizations from private sector sources, rather than government official statistical systems.

A significant challenge during the WTC drafting was the uncertainty of SDG indicators. Negotiations of the global SDG indicator framework were launched in 2015 after adoption of the goals under the leadership of the United Nations Inter-Agency and Expert Group on SDG Indicators ( IAEG-SDGs ). The IEAG led the development and standardization of 231 SDG indicators. This has resulted in 225 SDG targets having well-developed and internationally agreed methodologies to ensure comparability, accuracy, and reliability. However, as of 2023, many indicators lack consistent data reporting, particularly in several goals and regions(Eshetie, Reference Eshetie2022; United Nations, 2023). For example, less than half of the 193 countries have shared data for goals on climate action (SDG 13), gender equality (SDG 5), and strong institutions (SDG 16) (United Nations, 2023).

However, considerable variation remains in data availability between SDG indicators. A systematic literature review of SDG monitoring identified 100 datasets derived from big data, covering 15 goals, 51 targets, and 69 official indicators (Allen et al., Reference Allen, Smith, Rabiee and Dahmm2021). The largest share of papers corresponded to SDG 15 on life on land (21%), SDG 6 on clean water and sanitation (15%), SDG 1 on poverty (12%), and SDG 11 on sustainable cities (12%) (Allen et al., Reference Allen, Smith, Rabiee and Dahmm2021). The largest data gaps were found for SDG 13 on climate action and SDG 14 on life below water, for which few countries have the capacity to report data (Sachs et al., Reference Sachs, Kroll, Lafortune, Fuller and Woelm2021). These findings align with other analyses showing significant gaps in data availability for environmental targets (Dahmm, Reference Dahmm2021).

While the assumption of increasing data supply holds, three challenges need reconsideration. First, data supply does not always increase where needed most, with varied levels between SDG indicators and regions. Second, it is not only the supply but also the analytic methods that need clarification and standardization to leverage data value, particularly when using uncertainty science and synthetic data generation (Andree et al., Reference Andree, Mahler and Newhouse2023; Savage, Reference Savage2023). Finally, new data supplies do not always measure what matters most for the SDGs or inform key decisions, necessitating new approaches to prioritize data collection (Shepherd et al., Reference Shepherd, Hubbard, Fenton, Claxton, Luedeling and de Leeuw2015).

2.2. The SDGs would be the driving force for data innovations for public use (Assumption 2).

The second major assumption posits that the holistic, global, and integrated nature of the SDG framework would sustain political demand for data and statistical system innovations, particularly for national policy design and program implementation. This assumption builds on the experiences of the MDGs, where global coordination generated political attention and investments into national statistical systems (United Nations, 2016). The WTC authors acknowledged that MDG-framed initiatives successfully filled gaps in national statistical systems to track progress against country-level development plans, going further that the SDG momentum was expected to continue driving data generation, standardization, and use.

The SDGs have provided a political and financial mandate to advance pre-existing global commitments on data and statistics and driven UN activities. The Marrakech Action Plan for Statistics (MAPS), endorsed by the United Nations Statistics Commission in 2004, emphasized the role of official statistics in development monitoring, reinforcing the Monterrey Consensus, a 2002 global agreement on financing for development that highlighted the importance of data for accountability and policy effectiveness. Since 1999, PARIS21 has been strengthening national statistical agencies to bridge data gaps and, since 2015, has expanded its support by guiding countries in developing SDG-focused National Strategies for the Development of Statistics and tracking financial commitments through PRESS. Since 2015, UN SDSN Thematic Research Group on Data and Statistics, and cross-sector mobilizations by the Global Partnership for Sustainable Development Data, have driven strategic analysis and cooperation.

Numerous UN initiatives have driven and maintained collaborative data platforms to interpret indicator data for accountability and global progress assessments. However, there is no systematic assessment of these innovations or their contributions to SDG outcomes. Table 2 provides illustrative examples showing the breadth of international and national public use, but it cannot systematically assess this assumption.

Table 2.

Illustrative examples of SDG-driven data innovations and public uses

While the SDG agenda has spurred much collaboration and innovation internationally, progress on national data and statistical innovation has been much slower. Although some countries established central coordinating offices for national SDG implementation (see UNDP, 2017, p. 42 for eight country-specific initiatives), a series of cascading global crises has shifted policymaking away from deliberate multi-year approaches towards crisis response. For example, the COVID-19 pandemic forced governments to rapidly adapt data production to prioritize data informing near-real-time policy formation. Movement restrictions necessitated changes in data collection methods, shifting from face-to-face household surveys to mobile phone-based surveys, while social media digital transactions, and mobile money were used to track employment, food security, and social safety net distributions (Carletto et al., Reference Carletto, Chen, Kilic and Perucci2022; World Bank, 2020, 2021, p. 35).

National responses during COVID-19 drove a significant number of new data producer-user collaborations and non-official data approaches, which had been previously identified as part of the data revolution. For example, Jamaica established a national cross-sector hub combining data from 30 organizations (Young and Verhulst, Reference Young and Verhulst2017), while Sierra Leone collaborated with the UN Economic Commission to produce geospatial datasets. Finally, the pandemic accelerated open data initiatives and interactive data visualization platforms, such as the Johns Hopkins global COVID-19 dashboard (Koch, Reference Koch2021).

The assumption that the SDGs would be the main driver for innovation in public use of data held within the UN system is less clear at the international and national levels. Evidence suggests that the COVID-19 pandemic was a greater accelerant of the data revolution at national levels, while the SDGs provide an organizing framework for data innovation. The cumulative impact of the pandemic and the subsequent crises of inflation, global conflicts, and, most recently, the cutting of foreign aid, has reinforced a reactive approach to national policymaking rather than one driven by a holistic approach framed by the SDGs.

3.1. The SDGs would enable the financing resources needed for the data revolution to accelerate progress toward outcomes (Assumption 3)

The third assumption posits that the data revolution requires reversing the persistent underfunding of official statistical systems and generating innovative funding streams for both official systems and the wider data ecosystem. The WTC assumed that underfunding could be addressed by providing investment roadmaps, aligning global funding pledges with national statistical plans, and leveraging innovative financing mechanisms, including private sector participation. The WTC made three key recommendations: (1) develop statistical system plans with cost estimates; (2) advocate for funders to fulfill pledges; and (3) establish accountability processes to track funding commitments (IEAG, Reference Melamed2014).

Since the WTC report, efforts have been made to mobilize greater financing from internal domestic and international sources for official statistics, with progress across all three recommendations. However, funding remains significantly below target levels and has decreased in real terms. In this paper, our focus is primarily on international sources of funding for data and statistics, recognizing that future analysis will be possible when reporting on national funding is standardized and available.

Following the WTC report in 2015, SDSN TReNDS estimated that over US$1 billion per year of additional funding was needed for statistical and data systems to support and measure the SDGs, including an addition of US$200 million annually in international assistance to low-income countries (Espey et al., Reference Espey, Swanson, Badiee, Christensen, Fischer, Yetman, de Sherbinin, Levy, Chen, Qiu, Greenwall, Klein, Jutting, Jerven, Cameron and Milena2015). This funding would cover censuses, household surveys, agricultural surveys, geospatial data infrastructure, civil registration, vital statistics, administrative data, economic statistics, and environmental data.

Country-level statistical roadmaps with budget estimates followed (Swanson and Eele, Reference Swanson and Eele2016), alongside global estimates for gender-disaggregated data funding (Open Data Watch, 2021).

The 2017 Cape Town Global Action Plan for Sustainable Development Data (CTGAP) revised these estimates to identify US$5.6 billion annual cost for 75 low- and lower-middle-income countries and 69 upper-middle-income countries with $4.3 billion (77%) covered by domestic resources, leaving a $1.3 billion (23%) gap for external sources (Calleja & Rogerson, Reference Calleja and Rogerson2019).

This sets a target for a ratio of 0.7% of Official Development Assistance (ODA) to be directed towards statistical systems.

The 2018 Dubai Declaration, endorsed by the United Nations Statistical Commission in 2019, mandated a coordinated and demand-driven funding approach for national statistical systems, including public-private partnerships. In response, two multilateral initiatives were established: the World Bank’s Global Data Facility and the Complex Risk Analytics Fund (CRAF’d). The Global Data Facility mobilizes and coordinates donor support for data and statistics at various levels (World Bank, 2019), while CRAF’d pools funding for supporting fragile state partners to use data in addressing complex risks and post-conflict recovery (CRAF’d, 2023).

Progress in accountability and investment since 2015 includes tracking funding commitments. SDG indicator 17.18.3 tracks the funding levels and sources of national statistical plans under PARIS21’s custodianship. Three platforms now track investments and capacity development: Partner Report on Support Statistics (PRESS), the PARIS21 Statistical Capacity Monitor, and the World Bank Statistical Performance Indicators (Dang et al., Reference Dang, Pullinger, Serajuddin and Stacy2023; PARIS21, 2023c).

3.2. Information gaps are the primary reason for policy failure (Assumption 4)

The WTC framed the pathway towards improved outcomes as contingent on the information provided to the policy design and implementation process. The WTC authors assumed three elements for how the Data Revolution would deliver value to policy and decision systems: reducing data gaps would prevent policy failures, increasing data supply would improve program efficacy and ensure accountability at domestic and global scales, and enabling statistical agencies would drive whole-of-government data use to meet decision-makers’ needs.

In the subsequent years, limited systematic assessment has been made on the impact of improved data and statistics systems for SDG-related policy decision outcomes. UN Sustainable Development Progress reports consistently identify data production and reporting gaps as undermining policy making, not how data is used (United Nations, 2022, 2024). The 2024 Sustainable Development Report notes the considerable increases in data to monitor the SDGs with 51 per cent of indicators having more than two data points in more than half countries, however, it also notes that policymakers lack information to make timely, informed decisions due to significant variability in data availability between goals, countries and timeliness of data (United Nations, 2024). While there are a growing number of assessments around data production capacity and illustrative case studies of value for decisions, particularly at the urban level (Jain and Espey, Reference Jain and Espey2022), there are limited systematic comparisons between countries or impact on individual SDG goals.

The most comprehensive systematic review of this assumption is the global PARIS21 Statistical Capacity Monitor, launched in 2022, which assesses statistical agencies’ capacity. The 2017 Cape Town Action Plan for Sustainable Development Data (CTGAP) set measurable targets for statistical systems (World Bank, 2022). Within roughly 100 indicators identified developments in the capacities of national statistical systems, six focus on the use of statistics, of which one tracks the use of statistics in national policy documents, and the others focus on newspapers and use by international organizations (PARIS21, 2023c). While an important step forward, these indicators do not enable assessment of how data is being used, the data gaps needed for policy decisions, or the integration across decision-making processes.

As discussed in the previous section, experiences after 2015 also show that gaps in data disaggregation remain, from sex to age, and are often attributed as a driver of policy failures (ADB, 2021; Henninger et al., Reference Henninger, Swanson, Noe, Wahabzada, Pittman, Hadnot and Appel2023; Misunas et al., Reference Misunas, Cappa, Murray, You, Hattori, Beise, Krasevec, Carvajal, Hug, Evans, Petrowski, Murphy, Kumapley, Bain, Andrew Porth and Yu2017). One consistent illustration of these gaps is the limited disaggregation of data on women and girls, with 22% of gender-specific indicators producing reports regularly, thus inhibiting national efforts to monitor and achieve SDG 5 (Pryor and Seck, Reference Pryor and Seck2019). By 2023, just 30% of indicators for SDG 5 (gender equality) had adequate data (PARIS21, 2023b).

The WTC did not differentiate the data and statistical needs across different stages of the policymaking cycle, yet there is a range of assumptions, supported by wider policy analysis literature, on the value of data within this cycle (Davis et al., Reference Davis, Althaus and Bridgman2018). The difficulty of getting data used in policy formulation is that it is an inherently political non-linear process, arguably more so than the reporting or problem assessment stages. The WTC did recognize that the reality of policymaking is not neatly sequenced steps, but assumed that SDG policy was rationalist in centralized structures. Data analytics integrated into policy cycles informs, and thus potentially challenges, the decision-making process for resource allocation. Scholarship on policy approaches is drawing attention to how data and statistics are used in contested overlapping nodes of responsibility (Ostrom, Reference Ostrom2010), each seeking relevant analytic tools and with different hierarchies of relevance (Cairney, Reference Cairney2021).

In retrospect, the assumption that increasing statistical systems data supply would catalyze policy impact and implementation effectiveness has been challenged on several fronts.

First, many policy design decisions are driven by political interests, not technocratic ones. Efforts have been made to reimagine the SDG data ecosystem, shifting focus away from increasing supply towards sharpening the collection for what matters most to the policy challenge. Recent literature has discussed how decision makers are not asking the right questions to guide the collection of the data that matters most for their specific decision needs, and thus it is not a data production gap rather the lack of decision-relevant data (Cripps et al., Reference Cripps, Fischer, Santow, Afshar and Davis2023; Levy, Reference Levy2017; Shepherd et al., Reference Shepherd, Hubbard, Fenton, Claxton, Luedeling and de Leeuw2015; S. Verhulst et al., Reference Verhulst, Chafetz and Fischer2024). This implies changes to what data is collected, and the frequency needed to respond to decision needs. An example has been the change to definitions and measures for SDG 6 for global water, sanitation, and hygiene (WASH). The Joint Monitoring Programme has expanded the measures beyond access to include water safety and reliability, as well as measuring public institutions such as schools and health clinics (Chatterley et al., Reference Chatterley, Slaymaker, Badloe, Nouvellon, Bain and Johnston2018). But national surveys do not address real-time delivery, and new integrated water monitoring systems are responding to the decision needs of government agencies and water service providers simultaneously (Hope et al., Reference Hope, Foster, Koehler and Thomson2019; Thomson and Koehler, Reference Thomson and Koehler2016).

Second, while boosting the capacity of statistical agencies is a policy priority, assumptions about how to increase this capacity are constantly evolving. In recent years, there has been growing recognition that policymaking and implementation agencies often lack the internal skills to interpret and utilize this data effectively. Models like Germany’s data labs networked across government agencies offer examples of how to build this capability (Engler, Reference Engler2022). Although statistical agencies remain crucial for producing trusted official data, they operate within a broader data ecosystem that supports cross-sector decision making (Verhulst, Reference Verhulst2021). Collective intelligence frameworks, including accelerator labs and distributed data systems, are expanding the role of these agencies by focusing on the data and insights required for key decisions(Peach et al., Reference Peach, Berditchevskai, Mulgan, Lucarelli and Ebelshaeuser2021).

Third, as the WTC report anticipated, technological advances are transforming data systems alongside established statistical practices. However, while much of the data revolution has centered on boosting statistical capacity, this has not been matched by improvements in the capability of policymakers to use the data effectively. This suggests that the assumption should shift from focusing solely on reducing data gaps to addressing gaps in capability and processes within policy formation. Greater innovation is needed to bridge the interaction between politics and technocratic advances.

Finally, national statistical systems are assumed to be the most trusted and legitimate sources of data due to their rigorous methods, but the rise of post-truth politics has challenged the role of evidence in policymaking (Habermann and Louis, Reference Habermann and Louis2020). Public trust in statistical systems has eroded, exacerbated by the exclusion or misrepresentation of certain communities in datasets or the politicization of distrust of data systems (Pullinger, Reference Pullinger2020). This raises the challenge of requiring data as part of the policy design and accountability process.

The national policy ecosystem’s reliance on data maturity and analytic capacity underscores the need to balance expanding data production with building the capacity to use existing resources effectively for complex decision-making.

4.1. The public sector would drive and guide data innovations to target sustainable development (Assumption 5)

The WTC report placed national governments at the center of the data revolution, positioning them as both guarantors and drivers of innovation, contrasting with the assessment of internationally driven incentives during the MDG era (Florini and Pauli, Reference Florini and Pauli2018; IEAG, Reference Melamed2014). National governments were assumed to be the primary agents sustaining the data revolution by providing financial incentives, demand points, and pathways for applying data innovations. While the WTC report acknowledged the role of the private sector and civil society as key to data supply innovations, it emphasized government leadership in advancing SDG reporting, downscaling data to enable national policy needs, and guiding standardization (United Nations, 2015). The WTC also assumed that governments would oversee data protection, and drive innovation to ensure universal coverage in line with the SDG commitment to “leave no one behind.”

Building on the WTC recommendations and SDG mandate, the Global Partnership for Sustainable Development Data (GPSDD) was formed and has since built a network of 300 organizations with over 100 strategic partnerships and data collaborations with national statisticians and country-level SDG campaigners and data innovators (Melamed, Reference Melamed2021). This initiative was established to foster innovative practice and informal accountability for commitments.

However, since the SDGs were adopted, governments have been less central to data innovation than anticipated. While they continue to lead official statistical monitoring and Voluntary National Reviews, much of the data innovation has come from civil society, academia, and multilateral organizations, including citizen-generated data and geospatial applications (World Bank, 2021). Civil society has also driven partnerships for inclusive data policies, such as the Open Data Charter (Davies et al., Reference Davies, Walker, Rubinstein and Perini2019). Meanwhile, private sector data generators have sought government action, but government coordination of private-sector contributions remains limited (Li and Hinrichsen, Reference Li and Hinrichsen2023).

There are several areas where we see this divergence. The first is that many key data innovations and applications to the SDGs have been driven by civil society, academia, and multilateral organizations, including citizen-generated data production and geospatial data. (Fritz et al., Reference Fritz, See, Carlson, Haklay, Oliver, Fraisl, Mondardini, Brocklehurst, Shanley, Schade, Wehn, Abrate, Anstee, Arnold, Billot, Campbell, Espey, Gold, Hager and West2019; World Bank, 2021). The second is the role that civil society has played in driving new partnerships and coordination, motivated by the agenda for inclusive and open data to ensure no one is left behind, including contributions to policies such as the Open Data Charter and Inclusive Data Charter. (Badiee and Melamed, Reference Badiee and Melamed2014; Davies et al., Reference Davies, Walker, Rubinstein and Perini2019; GPSDD, 2022) The third is that governments have not coordinated or incentivized the private sector; rather, corporations and the private sector have sought government action to enable their contributions. (GPSDD, 2018; Li and Hinrichsen, Reference Li and Hinrichsen2023)

Several factors explain this shift. First, international efforts have prioritized building official statistical capacity for SDG monitoring over driving demand for innovative data (Besley et al., Reference Besley, Burgess, Khan and Xu2022; Andrews et al., Reference Andrews, Pritchett and Woolcock2017). Second, the decentralization of power to regional and municipal levels has led to more networked approaches to data use, leading to innovations from multiple actors rather than centralized government initiatives (Allen et al., Reference Allen, Malekpour and Mintrom2023). Third, much private sector data innovation has been driven by Environmental, Social, and Governance (ESG) reporting obligations, not government incentives. Until recently, companies adhered to voluntary reporting standards such as the Task Force for Climate-Related Financial Disclosures (TCFD) (Task Force on Climate-Related Financial Disclosures Guidance on Metrics, Targets, and Transition Plans, 2021), the Global Reporting Initiative, and the Sustainability Accounting Standard Board (SASB). Organizations like ESRI and Groupe Special Mobile Association (GSMA) have facilitated public-private collaboration, contributing to the growing ESG data landscape (ESRI, 2021). Opimas, a consultancy firm, estimates the global market for ESG data surpassed $ US$1.3 billion in 2022, a fivefold increase from 2015 levels (Foubert, Reference Foubert2022). However, despite substantial private investment, largely in high-income countries, coordination around a data revolution agenda remains limited.

This corresponds to recent scholarship and policy commentary on the emergence of collective intelligence, or knowledge systems driven by people, technology, and data, where simultaneous independent data inputs are pooled to create shared information systems. Collective intelligence systems are not driven by a single force; instead, they combine the incentives of multiple sources and actors (Mulgan, Reference Mulgan2017). Demonstrating steps towards this approach, the World Bank has recommended integrated national data systems (INDS) built around frameworks for whole-of-government and multistakeholder governance and legal frameworks for data protection and rights (World Bank, 2021) and the United Nations Development Programme has set up national collective intelligence data labs (Peach et al., Reference Peach, Berditchevskai, Mulgan, Lucarelli and Ebelshaeuser2021).

At the midpoint of Agenda 2030, it is clear that the private sector and civil society are playing a more significant role in driving data innovation for sustainable development data than anticipated. While the WTC envisioned a government-led data revolution, progress has come from collective intelligence and decentralized drivers. Opportunities exist for greater collaboration, such as aligning ESG standards with global statistical frameworks like the System of Environmental-Economic Accounting (SEEA), particularly for biodiversity reporting. Citizen-generated data, such as monitoring plastic marine litter, also holds potential for enhancing metrics that national statistical offices cannot capture.