To achieve the Sustainable Development Goals1 (Supplementary Table 1) and implement the Paris Agreement, developed and developing countries alike will need to transform their energy systems, ecosystem management, agriculture and land use, urban management, material use, gender outcomes, health, education, governance and other areas2,3. In addition to requiring greater financial resources and political commitments, these transformations will also place major demands on science to devise data and monitoring frameworks4, to relate planetary boundaries to national sustainability objectives5,6, to develop innovative solutions and to chart out integrated pathways for achieving the goals2,7, taking account of the trade-offs and synergies across goals and targets8,9.

The predecessors to the SDGs, the Millennium Development Goals (MDGs) that expired in 2015, mobilized attention on addressing the challenges of extreme poverty, hunger, illiteracy and disease10. The MDGs helped spur advances on many fronts. In health, the MDGs have been associated with a significant acceleration of progress in some of the poorest countries11,12,13, which stands in contrast to the lack of progress on environmental sustainability observed under the three Rio Conventions14 and other MDG priorities, such as access to water supply13.

The MDG experience suggests that global goals can serve as a management tool and report card that focus attention on complex sustainable development outcomes10 and accelerate progress towards these outcomes. Yet success is far from guaranteed. Inter alia, it will require educating decision makers and the public in sustainable development; mobilizing science for diagnosing challenges, identifying solutions, developing long-term pathways and tracking progress; mobilizing governments, businesses, and civil society for action around shared goals; and cooperation across countries to address planetary boundaries5 and other areas requiring international collaboration, such as implementing the Paris Agreement or aid-financed investments in developing countries.

Compared with the eight MDGs, which were extracted from the Millennium Declaration by a team of officials working under the former UN Secretary-General Kofi Annan10, the SDGs represent a political compromise negotiated by the 193 member states of the United Nations that has been critically reviewed9. In particular, the goals combine policy ends (such as ending extreme poverty or ending preventable child deaths) with means such as development finance and maintaining a global partnership for development. Many SDGs focus on flows instead of focusing on stocks, as recommended by many scientists15,16,17 since the report of the Brundtland Commission18. Finally, the goals do not propose a hierarchy among the 17 goals and associated targets. In this paper, we focus on how baselines for the SDGs can be established without aiming to resolve the criticisms of their design.

Good data and clear metrics are critical for each country to take stock of where it stands, devise pathways for achieving the goals and track progress. The UN Statistical Commission has recommended a first set of 230 global indicators to measure achievement of the SDGs, but many suggested indicators lack comprehensive, cross-country data and some even lack agreed statistical definitions19. More and better data are needed, but it will take years to build the necessary statistical systems even if adequate resources were mobilized, which is currently not the case20. Some governments have begun voluntary national reviews of progress on the SDGs, but they use indicators that are not harmonized internationally and lack comparability21.

In the meantime, scientifically robust tools are needed to help operationalize the SDGs at the global, regional, national and sub-national levels in order to begin a process of data-driven and evidence-based implementation and follow-up. Specifically, such tools should address the following questions: (i) What are the 2015 baselines for key SDG indicators at global, regional, national and sub-national levels? (ii) How far is a country from achieving a particular SDG, and which are the country's most important SDG challenges? (iii) How can countries' overall progress towards implementing the 2030 Agenda and the 17 SDGs be assessed, compared and tracked over time? (iv) Which data gaps need to be filled most urgently to support better monitoring and facilitate peer-learning between countries with regard to policies that help achieve the SDGs?

Reporting a large number of indicators alone, while providing much detail about specific domains, eventually leaves open the question of how to measure the aggregate performance of a country. Composite indices have well-known weaknesses22, but they can synthesize complex information into a single number and may be more effective in stimulating public debates than a large number of individual scores which could result in cherry picking15,20. To inform policies for the achievement of complex integrated goals, a combination of composite measures and dashboards are therefore needed for the SDGs. Prominent examples of such composite indices in recent years include the Global Burden of Disease index for the health SDGs23, the Environmental Performance Index (EPI)24, the Ocean Health Index25, the Human Development Index (HDI)26, the Programme for International Student Assessment education assessment27, the Global Competitiveness Index28, the Global Peace Index29, the Index of Economic Freedom30 and many others. Some of these composite indices have had significant impact in drawing political attention31,32, guiding policies, fostering learning across countries, and promoting more research33,34. Yet, only the health-related SDG Index by the Global Burden of Disease collaboration was specifically designed to track a part of the SDG agenda.

In order to assist countries in measuring their SDG baselines and to measure future progress, the Bertelsmann Stiftung and the Sustainable Development Solutions Network (SDSN) jointly released the first SDG Index and Dashboards in July 2016, with data covering 149 of 193 UN member states35. They build upon an SDG Index published in 2015 by the Bertelsmann Stiftung with the support of SDSN, covering the 34 developed countries that were members of the Organisation for Economic Co-operation and Development (OECD)36. In this paper, we present the SDG Index and Dashboards that will be updated and revised annually. We explore differences in countries' performance, consider the empirical relationship to subjective well-being (SWB), and discuss how remaining gaps in data and analysis can be filled.

The SDG Index and Dashboards

The annual SDG Index provides a standardized, quantitative, transparent and scalable composite measure of SDG baselines for 149 countries with sufficient data across the goals. It synthesizes 63 global indicators plus 14 additional indicators for OECD countries into an overall assessment of SDG baselines and ranks countries according to their starting points on the 17 SDGs. We included official SDG indicators19 with data available for at least 80% of countries with a population greater than 1 million. Indicator gaps were filled using published data from other sources (Supplementary Table 3). The authors consulted widely with expert communities on suitable indicators, including through a public consultation hosted by the SDSN. The methodology follows established principles from the academic37 and policy literature22. The details of the index are summarized in the methods summary and detailed in the Supplementary Information, which references all data sources and provides sensitivity analyses showing that the rankings are robust with regard to alternative specifications.

Table 1 shows the SDG Index ranking for a selection of countries. Sweden is ranked first, having covered 84.5 percent of the distance towards the optimum outcome across SDG metrics for which data was included in the index. The annual SDG Dashboards (Fig. 1) present baseline information by SDG to identify implementation priorities for each country.

Table 1 SDG Index 2016: score and ranking.
Figure 1: SDG Dashboards 2016.
figure 1

SDG Dashboard for members of the G20 and other countries with a population greater than 100 million. Dashboards for OECD countries calculated using an augmented set of 77 indicators, compared to 63 global indicators for non-OECD countries. Green signifies that the country has achieved the goal, yellow points to significant challenges that remain and red warns that major challenges must be overcome to meet the goal. Grey indicates an SDG for which there is no data. Icon images courtesy of United Nations.

Interpreting and applying the SDG Index and Dashboards

The SDG Dashboards show that even the wealthiest countries face major challenges in meeting several of the SDGs, confirming the universal relevance of the goals and the need to consider the full SDG agenda to avoid cherry picking. To meet the SDGs, poor countries must make substantial progress in ending extreme poverty, ensuring access to essential infrastructure, curbing environmental degradation and promoting social inclusion. Richer countries face fewer but nonetheless major challenges in addressing climate change, lowering inequalities, halting the loss of biodiversity and contributing their fair share towards the global partnership to achieve the SDGs.

The SDG Index is correlated with per capita gross domestic product, the most widely used indicator of economic progress; the HDI, a composite measure of health, education and income outcomes; the Global Competitiveness Index, a common index of countries' economic competitivenes; the EPI, which comprises a broad range of environmental indicators; and the Global Peace Index, a broad measure of peace and conflict (Fig. 2). The correlation is strongest with the HDI and GDP and weakest with the Index of Economic Freedom and the Global Peace Index.

Figure 2: Correlation of SDG Index with other common development indices.
figure 2

a–f, Pairwise, population-weighted correlation of SDG Index scores by geographic region and income group with: natural logarithm of GDP per capita (PPP)38 (a); Human Development Index26 (b); Global Competitiveness Index28 (c); Index of Economic Freedom30 (d); Environmental Performance Index (e); and Global Peace Index (f). Countries above trend line perform better on the SDG Index than suggested by correlations.

However, substantial differences exist in performance within regions and across regions and income groups (Table 2). For example, average SDG Index scores for East and South Asia are lower than estimated from the global bivariate relationships except for the EPI, suggesting that these countries have prioritized economic and social development over the environment. The reverse is true in countries in Eastern Europe and Central Asia, and Latin America and the Caribbean that perform better on the EPI than other regions. Sub-Saharan Africa performs worse on the SDG Index than on all other indices, except the HDI. This finding is consistent with significant investments in health and basic education under the MDGs that have yet to be matched by similar investments in other SDG priorities. This illustrates the usefulness of a broader SDG Index in identifying imbalances in countries' development needs.

Table 2 Relative performance on SDG Index by region and income group.

OECD members and high-income countries score better on the SDG Index than on the other indices suggesting that their development model is more balanced on average. However, as illustrated in the SDG Dashboards, OECD and high-income countries perform poorly on some goals, such as climate change, showing the limitations of only considering the aggregate SDG Index scores.

Tables 3 and 4 therefore disaggregate SDG Index scores by individual SDGs to identify which dimensions of the index drive each country's performance relative to others. Results show that different countries pursue different development models. For example, the United States ranks 9th in per capita GDP38, but 25th in the SDG Index. Relative to its overall SDG score, the country experiences major deficits in inequality and peace and justice (SDGs 10 and 16), environmental objectives (SDGs 12–15) and partnership for the Goals (SDG 17). Findings are similar for China, Russia and the UK. Meanwhile, continental European countries tend to be more equal, but face major challenges on the environment goals and in some instances on economic performance. Some countries, notably from the Middle-East and North Africa, perform well on meeting basic needs, as measured by the HDI, but fall short on the SDG Index. For example, Saudi Arabia ranks 35th in the HDI but 85th (both out of 149) in the SDG Index owing to deficiencies in gender and income inequality, and the environment. This shows how a combination of the SDG Index and the SDG Dashboards provide a richer understanding of a country's development challenges.

Table 3 Deviation from average country SDG Index score by SDGs 1–10.
Table 4 Deviation from average country SDG Index score by SDGs 11–17.

Supplementary Table 10 reports the SDG Index and its components for each country. This data can help countries identify when their development is imbalanced by benchmarking performance across individual goals with average country performance as well as performance of countries at a similar stage of development.

The SDG Index is also partially correlated with subjective well-being in the presence of common correlates considered in the literature — per capita GDP and unemployment (Box 1). This illustrates the usefulness of the SDG Index in understanding determinants of (and cross-country differences in) SWB and other policy objectives.

Major data gaps for the SDGs

Three types of data challenges need to be addressed to improve the measurement of baselines for the SDGs and to ensure effective monitoring, as summarized in Table 5. First, some SDG priorities lack scientifically robust indicator definitions that can be applied in a broad range of countries. Developing or improving such definitions will require a collaboration between the respective academic communities with statistical offices and policymakers. International organizations could convene such dialogues and support the production of handbooks on new measurement areas, such as the OECD Guidelines on Measuring Subjective Well-being39.

Table 5 Principal SDG data issues and gaps.

Second, some indicators require better and more frequent data collection and dissemination, particularly in poor countries and small-island economies where data is either unavailable or estimated too infrequently to allow for the reliable estimation of trends over time. In many instances this will require substantial additional investments in statistical systems and data collection mechanisms40. Countries with small populations need to consider alternative methods for approximating some data since standard survey techniques may become inoperable.

Third, in other areas data is collected by scientists or available through big data but is not adequately used to inform official SDG monitoring efforts at national, regional and global levels. This applies particularly to perception-based indicators (for example, on corruption or subjective well-being) but also other metrics, which are well-accepted scientifically, but not used by many national statistical offices and not included among the proposed official SDG indicators19. In some areas such as food loss and waste, greater efforts are needed to bring SDG-relevant commercial data into the public domain and to improve the availability of such data.

As a result of these data gaps some SDGs remain poorly measured. For example, we are unable to include indicators on gender-based violence. SDG 12 on sustainable consumption and production presents some of the greatest data challenges. This is true, for example, in the case of the Material Footprint Index41, which we did not include in the SDG Index for two reasons. First, with the exception of fossil fuels, which are covered under SDG 13, it is not clear how per-capita consumption of specific materials (biomass, construction minerals and metal ores) relates to local and global environmental impact. Second, the Material Footprint Index aggregates consumption across a broad range of different materials on a per-kilogram basis even though one kilogram of biomass might have a different environmental impact than one kilogram of iron ore or building stone.

Another important shortcoming in the current SDG Index and available SDG data is the inadequate measurement of spill-over effects of one country's actions on the ability of other countries to achieve the SDGs, a phenomenon sometimes referred to as leakage. Spill-overs include greenhouse gas emissions, land degradation, endangerment of species, fisheries depletion, forced/child labour, groundwater depletion, financial secrecy and other detrimental effects through global supply chains and international trade. Better addressing spill-over effects as well as sustainable consumption and production patterns in the SDG Index may significantly change the ranking of some high-income countries that for example consume large volumes of environmental resources or promote policies with negative impacts on other countries' SDG baselines. See Supplementary Section 4 for more details.

Implications for future research and policy

The SDG Index provides a first comprehensive assessment of countries' starting points on the internationally agreed SDGs. In contrast to GDP per capita and more narrowly defined indices, it addresses the full spectrum of economic, social and environmental challenges that countries face in achieving the SDGs providing additional insights into countries' sustainable development needs. The SDG Dashboards show significant variation in starting points across countries, and they underscore that every country falls short on a number of SDG priorities. Both tools show that countries pursue different development strategies, and that many countries need to better balance economic, social, and environmental objectives. These questions require further scientific analyses in all major disciplines relating to the 17 SDGs.

The scope and coverage of the SDG Index is currently limited by the availability of data, and significant gaps exist, such as sustainable consumption or gender-based violence. Official statistics do not adequately cover the goals, so they will need to be complemented by robust, scientific data. Future research into improving the SDG Index will focus on adding more SDG metrics as data becomes available and relating national performance thresholds to be achieved by 2030 to planetary boundaries5,6.

To better reflect regional priorities, the SDG Index and Dashboards can in the future be augmented with variables that are of particular relevance in a given region, such as malaria metrics in sub-Saharan Africa. Likewise, countries can apply SDG Indices at the sub-national level to compare starting points across states and provinces. A first prototype SDG Index has been launched for US cities42, and similar indices can be designed for cities around the world.

Finally, more countries need to be included in the Index. For example, compact versions of the SDG Index can be developed for small island developing states that tend to lack data for key socioeconomic and environmental variables owing to their small population size and limited data collection capacities.

In view of current data limitations, the SDG Index and Dashboards do not provide a comprehensive measure of sustainable development. The probable addition of indicators in future editions of the SDG Index, combined with the fact that data for many variables is only updated infrequently and with different periodicities, limits the scope for calculating year-on-year changes in the SDG Index and Dashboards and for using them as monitoring tools (see Supplementary Section 4). When publishing updates to the SDG Index we will consider how progress towards the goals can be estimated using the Index and Dashboards to help inform official SDG monitoring processes.


As described in Supplementary Section 1, the SDG Index and Dashboards include more than 230 official SDG indicators47 — proposed by the Inter-agency and Expert Group on SDG Indicators (IAEG-SDGs) — that met tests of data availability and usability in a global index. Where official SDG indicators did not meet the criteria for data selection or where indicator gaps remained, we considered official and other metrics published in the peer-reviewed literature, as well as major databases and reports on development and environmental indicators. We also consulted with a broad range of experts and conducted a public consultation on an earlier draft of the analysis that generated 56 submissions. Since one purpose of the SDG Index and Dashboards is to highlight missing data, we did not impute missing data except for four variables that lacked data for high- or low-income countries (Supplementary Section 1.3). Each of the 17 SDGs has at least 1 (SDGs 1 and 11) and a maximum of 11 indicators. The raw data for the SDG Index is available for download with this paper.

Data for each indicator was normalized on a linear scale of 0 to 100. A score of 0 was defined by performance at the 2.5 percentile to ensure extreme values did not skew the distribution22. A score of 100 denotes target achievement and “leaving no one behind”1 (for example, zero extreme poverty, 100% school completion). Some SDG target thresholds are below levels achieved today in high-performing countries. For example, the official 2030 SDG Target for child mortality is 25 per 1000 live births, whereas the top-performing countries have rates of 2.3–3.9 per 1000. For such SDG indicators, the maximum score of 100 was set as an aspirational target surpassing the SDG Target (for example, 0 child mortality per 1000 live births). In cases where no quantitative SDG target or an aspirational target could be identified we used the average of the top-five-performing countries as the benchmark for top performance. In this way the SDG Index defined an optimum outcome across the 17 goals that countries, including advanced economies, should aim for by 2030.

We aggregated indicators arithmetically within each goal and then averaged across goals, applying the same weight to every goal according to equation 1:

Where Ii is the index score for country i, Ni the number of SDGs for which the country has data, Nij the number of indicators for SDG j for which data is available for country i, and Iijk denotes the score of indicator k under SDG j for country i. This weighting for the SDG Index is subjective, as is the case with all composite indices22,37. Our approach is consistent with the intention of UN member states who framed the SDGs as an “integrated and indivisible” agenda1, whereby the goals have equal priority. Supplementary Section 2.3 discusses the motivation for and implications of alternative aggregation methodologies.

Our methodology differs in some aspects from the approach of the health-related SDG Index23 produced by the Global Burden of Disease consortium. Its authors relied on extensive modelling and interpolation to generate data for 188 countries48. Meanwhile, we refrained from modelling data due to the heterogeneity of the data sources used for the SDG Index, the absence of robust time series for some survey-based and other metrics, and the limited geographical coverage of many indicators for important SDG priorities. Moreover, the SDG Index seeks to draw attention to data gaps, so we limited the imputation of missing data to measures of extreme deprivation, which were not collected in rich countries and which we could confidently set at zero in the high-income countries (Supplementary Section 1.3).

For the health goals and several other SDGs, both the health-related SDG Index and the SDG Index define the upper bounds according to the average performance of the top countries. But this approach of relying on 'best performers' could not be used for greenhouse gas emission reductions, sustainable energy use or other areas where no country currently meets sustainability thresholds. For such variables, absolute target values had to be defined (Supplementary Section 2.2). The OECD pilot assessment used a similar methodology to ours and applied absolute thresholds (ref. 49), whereas in ref. 36 OECD countries are scored based on their relative performance. To construct the SDG Dashboards, each indicator was assigned absolute performance thresholds (Supplementary Table 6). For each goal a country's SDG Dashboard rating was determined by the rating of the worst-performing indicator. In this way, the SDG Dashboard highlights major implementation challenges in a goal even if the country performs well on several indicators. For example, a country that scores well on nutrition and food security indicators may be rated red on SDG 2 if it experiences high rates of obesity or low nitrogen use efficiency.

Data Availability Statement.

All data used for this study and results are available for download on Figshare (