How do Small Island Developing States Meet the Sustainable Development Goals?

,


Introduction
States in the Global South are facing a double challenge of achieving socio-economic development while adapting to climate change impacts. In this paper, we investigate to what extent Small Island Developing States (SIDS) manage to address this double challenge and meet the Sustainable Development Goals (SDGs). The SIDS consists of 58 marine territories (see Appendix, Table 1), including non-independent states, and was first recognized as a separate entity in 1992 at the United Nations Conference on Environment and Development. They are at the frontline of climate change tackling its consequences such as sea-level rise, ocean acidification, and increasing frequency of natural disasters (Betzold, 2015;Guillaumont, 2013;Scandurra, Romano, Ronghi, & Carfora, 2018;Thomas, Baptiste, Martyr-Koller, Pringle, & Rhiney, 2020).
The SIDS constitute a heterogenous group containing a great variety in terms of economic development, initial conditions, historical legacies, institutional structures, and factor endowments. Still, many SIDS share basic characteristics such as being remotely located and highly dependent on international trade, while having limited land-based resources, growing and often low-skilled populations, small domestic markets, and expensive public administrations (Becker, 2012;Briguglio, 1995;Guillaumont, 2013;Khor, Kronenberg, & Tumbarello, 2016).
These attributes complicate their ability to achieve economic, social, and environmental sustainable development.
Because their needs and capacity to respond to current challenges vary significantly, there is reason to expect their pathways to be versatile. The analysis of their varying development status is, however, muddled by the SIDS suffering from poor data coverage, meaning that data on their performance is scarce and scattered. Because of the poor data, SIDS are often cursory addressed in the ongoing international efforts to track and document sustainable developments in relation to the SDGs (Petzold & Magnan, 2019;Sachs, Kroll, Lafortune, Fuller, & Woelm, 2021). Consequently, while the SIDS work towards collaboration and building joint strategies, scholars, policymakers, and politicians within the SIDS as well as in the international community lack imperative information to analyse actual development levels, understand the current trajectories, and suggest alternative solutions.
In this paper, our overarching aim is to complement the existing broader international mapping and evaluation of SDG outcomes by highlighting SIDS specifically. By improving our understanding of SIDS's specific development status and challenges we hope to make the group more visible in the global debate and to contribute useful knowledge to the ongoing development work in and between the SIDS themselves. We pose three research questions: (i) How can we measure SIDS poor data coverage and what are the implications for our study? (ii) How do SIDS fare in relation to the three dimensions of sustainable development (economic, social, and environmental) embodied in the 2030 Agenda? (iii) What are the similarities and differences between SIDS in meeting the SDGs?
Our study confirms that SIDS are highly overrepresented among the most data poor countries in the world. By default, studies of SIDS relying on statistical data to evaluate SDG outcomes (or other development targets), including the present one, experience substantial limitations regarding what and who can be analysed. This finding emphasises the dire need to improve the SIDS's statistical capacity.
Further, we evaluate the SIDS's development performance, defined as meeting the SDGs with the three dimensions of sustainability (economic, social, and environmental), compared to a constructed Global Average (GA) excluding the OECD 1 . While SIDS overall resemble other developing countries, there are a few aspects where they deviate. First, there is a positive result as poverty levels measured by international standards are generally relatively low. Further, there are two adverse results that stand out -health hazards due to high adult obesity levels and low gender equality levels, especially manifested in the share of women in parliament. Finally, our results give cause to further reflect on SIDS' potential to develop alternative Blue Economy-led trajectories based on their large ocean territories and substantial marine resources. However, for more insightful policy recommendations for individual SIDS, our cursory overview needs to be complemented with additional country specific research.
The rest of the paper is structured as follows. In section 2, we outline the development of the SDG agenda and its relevance to the SIDS. Then we discuss the SIDS' development challenges more specifically. Sections four and five contain our analysis. First, we map the (poor) data availability, select our case countries, and motivate our choice of SDG indicators. Next, we compare the selected SIDS development performances to a Global Average (GA) that we construct excluding the OECD countries. We discuss our results for the three dimensions of the sustainability agenda (economic, social, and environmental), including potential patterns and identification of country outliers. In the final section, we conclude.

Defining and Evaluating Sustainable Development
In 1972, the Brundtland Commission Report introduced sustainable development as 'development that meets the needs of the current generation without compromising the ability of future generations to meet their own needs'. The concept has become the leading development paradigm within the global policy agenda, and in 2015 the UN launched Agenda 2030 with the call to developing and developed countries alike to incorporate a broader understanding of the well-being of people and planet. The ambition is to work towards positive economic and social impacts while at the same time confronting the environmental challenges of climate change and extinction of species (Mensah, 2019;Taylor, 2016). The Agenda 2030 is translated into 17 Sustainable Development Goals (see Figure 1) with a list of 169 targets and 232 indicators. The goals are interdependent and interlinked, bringing to the fore synergies and trade-offs that nations and policymakers need to prioritize based on country-specific needs and resource availability (Breuer, Janetschek, & Malerba, 2019. While Agenda 2030 is guiding much of the current work of international organizations, there is also substantial criticism against the content, measurement, and ambition of the SDGs. As climate change is progressing with increasing speed, some argue that they are 'too little, too late' and that a much more radical agenda is needed (Kotzé, 2018). Others point to the impossibility of the ambition to cater for all aspects of development at the same time. For example, Hickel (2019) points out that historical experiences show that it is not feasible to expect a decoupling between the growth process and material footprints such as reductions in resource use and CO2 emissions. The balancing-act between economic development and ecological impact becomes even more complicated when considering the Leaving No One Behind (LNOB) principle (Klasen & Fleurbaey, 2019;Nilsson, Griggs, & Visbeck, 2016). Within rich countries, there is an opportunity to address LNOB primarily through the redistribution of resources and a strive for reduced inequalities. Such strategies relate to sustainability ambitions presented within the literature on de-growth and circular economy, arguing for a need to think differently about our consumption levels and patterns (Latouche, 2009;Korhonen, Noor, Feldmann, & Birkie, 2018). Meanwhile, for the developing countries, improved government finances paying for social development and material improvements for the wider population need to be accompanied by economic growth. Finally, some researchers question whether the two big assumptions on which the SDGs rest, sustained economic growth and globalization, are still valid in the aftermath of the Covid-19 pandemic (Naidoo & Fisher, 2020).
While most of the current global development efforts stay focused on the SDGs, the monitoring of global goals as well as the task of measuring and evaluating progress for individual societies at very different stages in their development, is an elusive task. Since the implementation of the SDGs in 2016, researchers Jeffrey Sachs, Guido Schmidt-Traub, Christian Kroll, and Guillame Lafortune have published The Sustainable Development Goals Report. In the 2021 issue, they gave the most recent presentation with statistics for 211 indicators (Sachs et al. 2021: 5). These indicators were selected for: 1) monitoring the achievements of the SDGs; 2) being internationally comparable; 3) representing valid and reliable measures; 4) being up-to-date; 5) representing the best available measures; and 6) covering at least 80 percent of the 149 UN member states with a population greater than 1 million (Lafortune, Fuller, Moreno, Schmidt-Traub, & Kroll, 2018: 7-8).
The report is recognised as an important instrument for the research and policy communities to evaluate individual countries' and regions' progress in relation to Agenda 2030. However, it also reveals the challenges for developing countries, including the SIDS, to have the statistical capacity to deliver comparable data and thereby be included in the global development agenda. This means double challenges. First, while Agenda 2030 adhere to the LNOB principle, countries experiencing data poverty and being unable to measure and document their status in relation to the SDG targets and indicators risk being excluded from the global dialogues and joint efforts. Second, states that lack sufficient data to determine their own development trajectory will grope in the dark when determining the current status, evaluating possible progress, and setting up new targets for the road ahead.
Since the establishment of the SIDS group and the start of the territories' cooperation, the lack of data has repeatedly been recognised as one of the main vulnerabilities preventing the members to have a proactive response to, for example, the threats of climate change, food insecurity, overdependence on trade, and political legitimacy. Data management capacity was identified as a priority at the Barbados Programme of Action, BPOA, in 1994, and it was again called upon at the Mauritius Strategy in 2005 and the Third International Conference on SIDS organized by the UN in 2014 in Samoa. Pre-dating the Agenda 2030, these conferences aimed to commit, cooperate and implement shared strategies and solutions in the fight against climate change and increased frequency of natural disasters while ensuring socio-economic development and environmental conservation.

Challenges for Small Island Developing States
One of the key advantages of the current sustainable development agenda is also a core challenge -it is global and expected to be applied by almost 200 countries with very diverse contexts and development trajectories. While they constitute a heterogeneous group, many SIDS still share similar hindrances to sustainable development such as: remoteness; small landmass and under-exploited marine resources; limited human capital in terms of both size of population and levels of education; high communication, energy, and transportation costs; and excessive dependence on international trade. These characteristics combined with vulnerability to environmental shocks, including natural disasters and climate change, exacerbate the challenges of reaching the SDGs (Guillaumont, 2013;Scandurra et al., 2018;Thomas et al., 2020). While the literature on the SIDS remains limited, it contains some relevant considerations and discussions.
A first concern when analysing SIDS is their geographic location. While the Caribbean SIDS are located close to other large economies, the Pacific Islands suffer from remoteness, which increases the costs of transportation and communication with the international economy (Borgatti, 2008;UNCTAD, 2014). In addition, some Pacific Islands are archipelagic and their spatial fragmentation carries costs also within the nations themselves. Even with small and mostly homogenous populations, spatial fragmentation can be a challenge to internal politics and social cohesion in the long run.
There is a common stylised expectation that large states are richer than small ones because they have more natural resources as well as larger populations and domestic markets. However, many small states have proven that they are economically successful, and Easterly and Kraay (2000) have shown that small states in general do not fare worse than large economies. Their results are explained by small states compensating for their limited domestic markets by engaging in international trade. This is one important reason why SIDS tend to have a high degree of openness (Auty, 2019).
'Small states' are defined based on their limited land resources, including fertile arable land. However, when including the size of their ocean territories, many SIDS become sizeable and when the land and sea economies are integrated, the potential benefits that a country can derive from the marine resources contained within its exclusive economic zones compensate for the inefficiencies resulting from (landmass) smallness and remoteness (Khor et al., 2016).
Ocean resources and coastal areas are the basis of the Blue Economy (BE), a concept that aims to move beyond business as usual and to consider economic development and ocean health as compatible propositions. It promotes economic growth and social inclusion through the preservation or improvement of marine-based livelihoods while also ensuring environmental sustainability (Qi, 2022;World Bank, 2017).
Moving on, research shows that climate change has intensified the risks associated with natural disasters (Banholzer, Kossin, & Donner, 2014). For small states, sea-level rise is the main concern because it destroys the coastal infrastructure as well as the livelihoods and long-term habitability of islanders who then may seek relocation or outward migration as a solution. Solutions for adaptation to climate change 2 are costly and are estimated to take up from two to five percent of GDP (Robinson & Dornan, 2017). The toll on the budget is high, especially for small island economies that have relatively high levels of debt and expensive public administrations. While government policies do make a difference in the way SIDS deal with risks associated with climatic shocks (Petzold & Magnan, 2019;Sjöstedt & Povitkina, 2017), the final solutions to this environmental challenge usually lie in the hands of the large emitters rather than in the SIDS (Mackey et al., 2019).
Energy importing has set institutional barriers to decoupling from fossil fuels. The SIDS-DOCK 3 is an initiative among member countries of the Alliance of Small Island States (AOSIS) to help SIDS transform their energy sectors and address adaptation to climate change. A recent study examining the relationship between climate funding and energy substitution included some SIDS as recipients and concluded that several of them 4 have made significant improvements relative to other developing countries (Scandurra, Thomas, Pessaro, Bencini, & Carfora, 2020).
Finally, many SIDS struggle with improving the quality of their human capital in terms of health and education, and they are sensitive to outmigration (Kelman, 2015). Apart from the effects of climate change on agriculture, food imports that undermine local food production led to food insecurity and import dependency. Furthermore, food price inflation harms the nutrition, health, and living standards of low-income groups (Connell, Lowitt, Saint Ville, & Hickey, 2020).
Better human development comes with gender equality, and the effort is tilted toward the recognition of women's empowerment, their access to education, including the notion of education for sustainable development, and the realization of reproductive rights (Crossley & Sprague, 2014;Kelman, 2015). Meanwhile, the pace of migration by women and youth has accelerated, with education as the main factor behind this change (Bernard & Bell, 2018). Not surprisingly, recent research indicates that the political voice and labour of women and girls are key to having a transformative and multiplier effect on sustainable development in SIDS (Ramtohul, 2020).

Data Sources, Quality, and Coverage
While the importance of data collection, management, and disclosure has been emphasised at the three international conferences on SIDS - Barbados (1994), Mauritius (2005), and Samoa (2014) -the challenge remains. To say something of the magnitude of the problem, we make a data inventory of the SDG indicators from the United Nations for all countries in the world. We assign points per available indicator to assess the data availability: two points per year for data available after 2010, and 1one point per year for data before 2010. We allocate more points for more recent data, as it is more valuable for the task of measuring the current state of sustainable development (for results see Appendix, Table 2).
We find that the United Kingdom gets the highest score, 1005 points, closely followed by other Western European countries. Meanwhile, the majority of SIDS receives scores of 200 to 500 points. Subsequently, we divide the countries into three groups: 1) good data coverage; 2) medium data coverage; and 3) poor data coverage. Most SIDS end up in group three with poor data coverage, although the Dominican Republic, Mauritius, and Haiti are in group two. Within each group, countries are sorted from the highest to the lowest points. Our ranking shows large differences between the SIDS, from the high-ranking Dominican Republic with a score of 738 to the lowest scoring islands St. Kitts and Nevis (265), Tuvalu (264), and Nauru (216). Except for a few micronations (Andorra, Monaco, Lichtenstein, and San Marino), the bottom ten countries with poor data coverage in the world are SIDS. If we also exclude South Sudan, Equatorial Guinea, and the Democratic Republic of Korea, all of which are countries with substantially poor state capacity, SIDS make up the twenty most data poor countries in our global comparison.
The outcome for SIDS in the data coverage ranking exercise supports the argument that SIDS generally are struggling to collect the basic statistics to receive proper recognition in the global sustainability debate. The point can also be made by counting research publications. For example, adaptation research, which is central for understanding and reducing the costs of climate change and natural disasters, has been studied only in 26 of the 58 SIDS. Fiji received the most attention, with 12 articles examining the performance of adaptation solutions, while Jamaica had five articles and Kiribati, Samoa, and the Solomon Islands appear in four articles each (Klöck & Nunn, 2019).
For the present study, we identify our cases based on the selection criteria to only include independent island states with a minimum of data coverage for the years 2000-2018 available from United Nations (SDG dashboard 5 ) and World Bank data (World Development Indicators/ World Governance Indicators). This leaves us with 32 of the 58 SIDS affiliated with the United Nations in the last two decades. Henceforth when we discuss the SIDS in our analysis, we refer to this group of 32 purposefully selected countries.
Despite having weeded out the SIDS with the poorest data, we lack full data coverage for our remaining countries. It is primarily the smaller states that have gaps in the data, and the largest omissions are for SDG 12 Responsible consumption and production and government efficiency used as an indicator of SDG 16 Peace and justice and strong institutions. We adjust to these data limitations when deciding on indicators on which to base our analysis of levels of development 6 . We address SDGs 1-16 but exclude goal 17 as we deem the assessment of strengthening the means of implementation and revitalizing the global partnership for sustainable development to be too multidimensional for our approach. We select 1-4 indicators per SDG based on both relevance and data availability (for a summary see Appendix, Table 3).
The data points for each SIDS give an average of ten years (2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020). This allows us to include more countries than only those with full data coverage over all years. Moreover, studying ten year averages instead of one data point per country provides a fuller picture of how each of the SIDS fare over time. Still, we remain with some important limitations. First, we impose limited precision in our analysis as we do not compare the exact same years for all countries. Second, lack of data in significant areas, for example for SDG 12 Responsible consumption and production, means that we reach conclusions about SIDS's overall performance based on particular goals or indicators only. Still, we make a contribution as our analysis is one of the few studies addressing the Agenda 2030 ambitions with a clear SIDS focus (Jabbari, Motlagh, Ashrafi, & Abdoli, 2019;Van Beynen, Akiwumi, & Van Beynen, 2018).

SDG Outcomes in a Comparative Perspective
Our selected SIDS make up a heterogeneous group of countries in terms of GDP per capita levels. According to the World Bank, Haiti and the Comoros are classified as Lower-Income Countries (LIC), the rest are mostly classified as either Low-Middle-Income Countries (LMIC), or Upper-Middle-Income Countries (UMIC), but there are also several High-Income Countries (HIC), especially in the Caribbean. Although the six HICs had a GNI per capita income of US$ 12 696 or more in 2020, there remains a large difference between them and the OECD average of US$ 45 007 in the same year (see Table 1). Due to the substantial difference in socio-economic development levels between the SIDS and the OECD countries and because certain relevant development indicators such as access to clean water and sanitation are not measured for OECD, we compare the SIDS to a constructed Global Average (GA) excluding the OECD. To present our results, we provide tables summarising our values. To enable regional comparisons, we continuously present our 32 SIDS by ocean according to the three UN geographical areas: (i) the Caribbean, (ii) the Pacific, and (iii) the Atlantic, Indian Ocean, and South China Sea (AIS). We fill cells with red if the SIDS' average value is above the GA, and green if it is below. The colour is lighter if the value is within 10 percent above or below the average value of the comparison group. The grey cells indicate that we lack data. For a fuller analysis, we have also conducted comparisons with the other LICs, LMICs, UMICs, and HICs in the world. When relevant, we refer in the text to these additional comparisons that are compiled in the Appendix, Table 4 A-C.
We divide up the analysis into three sub-sections in line with the three dimensions of the sustainable development agenda -(i) Economic -SDGs 8, 9, 10, and 12; (ii) Social -SDGs 1, 2, 3, 4, 5, 7, 11 and 16; and (iii) Environmental -SDGs 6, 13, 14, and 15. In each sub-section we discuss the results in our tables, goal by goal.  There is no pattern indicating that the national income levels are correlated to size in landmass or population. For example, the small islands of Nauru and Tuvalu with an area of close to 25 square kilometres have a better performance in income per capita than many other SIDS.

The Economic Dimension
The results regarding the relationship between income levels and poor data coverage are mostly as can be expected, but also include some discrepancies. Our ranking in Appendix, Table 2 shows that the small Pacific Island nations have the poorest data availability, and generally their GDP per capita levels are below the GA. The exception is Palau which is data poor but classified as a HIC and has a GDP per capita of almost US$ 11 000. St Kitts and Nevis and Seychelles are other examples of HICs that are data poor. Meanwhile, Haiti, one of our LICs having the second to lowest GDP per capita (US$ 1193), is a country with medium data coverage together with Mauritius, a HIC with US$ 7897 per capita. The Dominican Republic has the highest data ranking, although it is a UMIC well below the GA income.
Moving on to industry, innovation and infrastructure (SDG 9), we can conclude that except for Trinidad and Tobago, the SIDS have low industry value-added, measured as a share of GDP. While SIDS generally do not have an abundance and wealth of land-based natural resources, Trinidad and Tobago has a sizeable petroleum industry, and oil and gas production represented an average of over 60 percent of exports between 1994 and 2018 (GrowthLab, 2022). The results showing the share of the labour force employed in industry are quite mixed with many SIDS close to the GA of 18.7 percent. Unsurprisingly, Trinidad and Tobago is above at 29 percent, but both Mauritius and Tonga also have above GA shares. In the case of Mauritius, this is due to a labour-intensive industrialization and economic upgrading trajectory spearheaded by the long-standing garment industry (Tang, 2019). At the other end of the scale, we find Haiti at 7 percent and Vanuatu at a bare 5.4 percent of the labour force in industry.
Infrastructure is particularly important for the SIDS economies for several reasons. For example, it provides the technologies to overcome remoteness, which functions as a protective tariff on the economy when communication and transportation are high (Pratt, 2015). Also, as the number of disasters increase internet access becomes an important channel to inform and manage the disaster risk reduction plans assigned in most national development plans (Jerez Columbié, 2022;Mackay et al., 2019). While the expansion of internet use has been pervasive, the differences across SIDS reveal that it is higher in the richer Caribbean and the AIS countries than in the Pacific.
For SDG 10 Reduced inequalities, we use the reported GINI as our indicator. The GINI values for all SIDS are rather compressed and most of them for which we have data are found in the medium range compared to the GA. While the GINI measures the relationship between different income groups, it does not tell us anything about poverty reduction or if the overall standard of living is improving for most of the population. At the country level, we see that SIDS in the Caribbean, in general, have a higher GINI than countries in the other two oceanic areas. These countries are also amongst the richest and there are large surpluses that can be unevenly distributed. St Lucia, a HIC in the Caribbean, has the highest GINI of all measured SIDS with an average value of 51.2. Meanwhile, Timor-Leste in the Pacific, a LIC, has the lowest average Gini, 30.8.
Most SIDS have relatively small populations and generally rely on tourism, financial services, external remittances, or the export of natural resources as key engines for the economy. It is not surprising then that, with the exceptions of Trinidad and Tobago, Jamaica, and Palau, SDG 12 Responsible consumption and production measured by adjusted savings in carbon emissions is below the GA. The high result for Trinidad and Tobago is explained by the petroleum industry. The results are more surprising for Jamaica and Palau which have different comparative advantages as Jamaica´s grow trade flows consists to 44 percent of travel and tourism, while Palau´s export relies on fish (60 percent), floating structures for scrapping (10 percent), and computers (10 percent) (GrowthLab, 2022).

The Social Dimension
Our presentation of results for the social dimension starts with a discussion on the three measures of poverty that we include in Table 3 below. The World Bank defines living on less than US$ 1.90 per day (2011 PPP) as living in extreme poverty. Eight SIDS spread across our three ocean groups have a share of their populations living in extreme poverty that is higher than the 17 percent that is the GA. In the Caribbean, Haiti has a share that is 23 percent. Among the AIS countries, Cabo Verde, Comoros, and Sao Tome, and Principe stand out as having substantially higher shares of the population living on less than US$ 1.90 a day, compared to the other countries in the AIS group. Four counties in the Pacific (Fed. Sts. of Micronesia, Papua New Guinea, Solomon Islands, and Timor-Leste) have rather high shares of the population in extreme poverty with Papua New Guinea having the highest share of all SIDS with 34 percent of the population living on less than US$ 1.90 a day.
We move on to our second poverty line, which is living in moderate poverty on less than US$ 3.20 per day. The same eight countries that had higher rates than GA in terms of extreme poverty also have higher levels of moderate poverty than the GA, but they are now joined by Vanuatu and Kiribati in the Pacific. Papua New Guinea is still the country with the highest poverty numbers with 65.6 percent of the population living in moderate poverty. poverty. For this measure, more SIDS are above the GA. Although the countries in the Caribbean in general have a lower share, and are often below the GA. The only Caribbean country with a share of the population living on less than US$ 5.50 per day that is above the GA is Haiti at 78.6 percent. More than half of our Pacific countries have shares that are higher than the GA. Timor-Leste, Solomon Islands, and Papua New Guinea have a share that is above 80 percent.
For SDG2 Zero hunger we display two indicators: percent of stunting (low height for age under the age of five) and percent of wasting (too thin for height). Stunting and wasting are measures capturing malnutrition and/or starvation in a population and we are expecting rather low levels of this in the SIDS as only two of them belong to the LIC group. The highest levels of stunting are found in the Pacific, with Papua New Guinea and Timor-Leste having stunting levels of around 50 percent. In the AIS group, Comoros has a stunting level of 32 percent and in the Caribbean Haiti is the only country with levels of stunting that is above GA. For wasting, we see similar patterns with Papua New Guinea and Timor-Leste having the highest prevalence while other SIDS are doing quite well.
For SDG 3 Good health and well-being, most SIDS have a life expectancy that is above the GA of 66.8 years, although some of the poorer SIDS have somewhat lower life expectancy. Furthermore, most SIDS are performing well in terms of child and maternal mortality. However, for both these two measures certain countries stand out.
In the case of maternal mortality, Haiti, Comoros, and Cabo Verde all have maternal mortality rates that are substantially above the GA and for under-five mortality, we see rates above the GA for Haiti in the Caribbean, the AIS countries Sao Tome and Principe and Comoros, and for Kiribati, Papua New Guinea, and Timor-Leste in the Pacific.
Meanwhile, obesity in the adult population is a major issue for many SIDS as their values tend to be higher than the GA. This is true for most countries, but particularly for some of the Pacific Islands. For example, in Nauru and Palau, over 50 percent of the adult population is obese, and in the Marshall Islands, the prevalence is over 48 percent. The obesity problem of Pacific Island nations is a well-known public health concern (Sobers & Samuels, 2019). What is striking in our analysis is that we see obesity levels that stand out also in the Caribbean, with most countries (except Trinidad and Tobago) having obesity rates that are higher than the GA. Meanwhile, all AIS countries have rates below the GA in this indicator. When comparing the obesity rates of the SIDS (in particular the Pacific Islands) to the LICs, the difference is striking, and it is upheld also in relation to the LMIC (see Appendix, Table 4 B). The obesity rate for LICs is 5.69 percent and all SIDS, except the Maldives and Timor-Leste, have higher rates than the average LICs.
The final measurement is vaccine coverage, denoted by the share of infants receiving WHO vaccines. For this indicator, there is a clear difference between the different oceans, as the lowest vaccine coverage is found in the Pacific where most countries also have coverage that is below the GA.
Moving on to SDG 4 Quality education, measured by mean years of schooling, we learn that the SIDS are doing rather well. SIDS in the Caribbean, excluding Haiti and Sao Tome and Principe, have the highest average years of schooling. The SIDS in the AIS still have a way to go to improve their results to reach the GA, except for Mauritius and Seychelles that are also the richest countries in the group. The SIDS in the Pacific do well, but their poorest members -Papua New Guinea, Solomon Island, and Timor-Leste -have a mean below the GA.
The next area of interest is gender equality (SDG 5). Female labour force participation, measured as a share of male labour participation, is below the GA of 67.9 percent in 12 countries in the sample. In the countries below average, it ranges from 48 percent in Timor-Leste to 67.7 percent in Comoros. The highest number for the countries above the GA is Papua New Guinea where female labour force participation is 96.4 percent or almost equal to that of their male counterparts. In terms of women in national parliament, the SIDS generally score under the GA of 15.3 percent, but the span is wide. In the Federal States of Micronesia, it appears that there are no women in parliament at all which is striking, while in Cuba, the share of women in parliament is 40.4 percent. The overall result that the SIDS score poorly in gender equality stays more or less the same independently of the income group we compare with. Turning to affordable and clean energy (SDG 7), it is recognised in the literature that SIDS generally, and the Pacific specially, suffer from energy poverty (Dornan, 2014). In addition, we show clear differences between our three geographical areas regarding access to clean energy. Compared with GA, the countries of the Caribbean have more access to clean fuels than their counterparts in the Pacific, where most countries have a lower mean than the GA.
Regarding peace, justice, and strong institutions (SDG 16), we see that the Corruption Perception Index is rather high in most of the SIDS, although there are differences within the group. For example, Barbados scores 70.8 in the index, while Haiti has a value of only 19.3. This is a somewhat puzzling result considering what we know about the overall state capacity of the two countries. We added data from the World Bank on Rule of Law. It captures the extent to which agents have confidence in and abide by the rules of society. The World Bank data consist of multiple indicators on e.g. property rights, crime, and law and order. The GA of this indicator is 37.2 on a scale of 100 and the majority of SIDS is above the GA.

The Environmental Dimension
The third and final part of the analysis is the environmental dimension (see Table 4). First, access to basic water and sanitation (SDG 6) shows overall positive results, although SIDS are doing better with access to clean drinking water than with sanitation. While the achievements are mixed, they to some extent correlate with the income levels of the different countries. Countries that are below GA in access to sanitation are all classified as LICs or LMICs. The exception is Samoa, which is a LMIC but with the highest score of all providing sanitation to 97.3 percent of its population. Regarding drinking water, almost half or 14 countries provide it to well above 90 percent of their populations, and Mauritius and Tonga score above 99 percent. The Caribbean SIDS generally tend to have better access to drinking water than the countries of the Pacific, and only one country in the Caribbean, Haiti which is a LIC with 45 percent access rate, is below the GA. Meanwhile, four LMIC countries in the Pacific (Kiribati, Papua New Guinea, Solomon Islands, and Timor-Leste) have lower access rates than the GA.
The SIDS turn out rather well in terms of SDG 13 Climate action, which we measure based on CO 2 emissions from energy. Considering that SIDS are developing countries with low levels of industrialisation, this is an expected  Vol. 16, No. 1; result. If we compare the SIDS with HICs, only two countries, Trinidad and Tobago in the Caribbean and Palau in the Pacific, have values higher than the HIC average of 10.6 tCo2/Capita. Because of its petroleum industry, Trinidad and Tobago has exceptionally high emissions and stand out with 33.8 tCo2/Capita.
All SIDS are highly dependent on their marine resources and most of them, especially the Pacific Islands, have large ocean territories. It is thus an expected result that these island states come out well in terms of SDG 14 Life below water when compared to the GA which includes countries that have short shorelines or are even land-locked. The Pacific countries moreover have higher values in terms of Ocean health /Fisheries, which also might be related to the larger marine areas of the Pacific compared to the Caribbean in particular. The future sustainable exploitation and preservation of the marine environment, the Blue Economy, constitutes an enormous opportunity and challenge to the SIDS.
Finally, we have previously alluded to the small landmass of the islands, and it is therefore not surprising that SIDS come out primarily below the GA for terrestrial sites and mean protected areas as well as in the Red list index of land species survival (SDG 15). There are six countries spread over the three oceans 9 that score better than GA, but it is difficult to find specific or principal explanations for their results. The overall result of scoring poorly remains strong independent of what group of countries we compare with.

Conclusion
The SIDS' collaboration towards a sustainable future can be traced back to the group's establishment 30 years ago and pre-dates the Agenda 2030 efforts. Due to the recognition by the UN as an entity and by virtue of standing at the frontline experiencing the repercussions of climate change, they receive relatively substantial attention from the international community despite generally being small states in terms of population size and landmass. From the onset, the group has faced the challenge of finding common ground while being a highly heterogeneous group.
In this study, we contribute new knowledge by mapping SIDS' performance in relation to the SDGs, but more research is needed to understand better the commonalities and differences of SIDS on which to base country specific development policies.
We first investigated the data availability. Collecting official statistics is a costly exercise and requires countries to have the infrastructure and capacities to produce statistics that are reliable and of high quality. Of the 32 independent island states included in our study, 28 were categorised as 'poor data coverage' in our ranking exercise. Further, together with a few failed states, SIDS make up the twenty most data poor countries in the world. We both confirmed an existing perception and provided new insights into the magnitude of the data poverty problem. We adjusted to this challenge by studying averages over ten years, and by being flexible when selecting the indicators per SDG, which we organised according to the three dimensions of sustainable development -economic, social, and environmental.
The SIDS are by definition developing countries and it is therefore expected that they are flawed in the economic dimension. However, the range of incomes is wide from the LICs Haiti and Comoros to HICs such as The Bahamas with almost US$ 30 000 per capita (US$ 2010) and Mauritius which is considered one of the African growth miracles. More in-depth studies could tease out what the poorer SIDS can learn in terms of opportunities for productive activities from their HIC counterparts. Also, with the exception of Trinidad and Tobago, SIDS have low levels of industrialization, but a sustainable future requires all economies in the world to renounce brown industries and instead develop so called 'industry without smokestacks', and to re-define what are 'industrial jobs'. The SIDS who are not burdened with brown industry can instead aim for green industrialization while preserving their positive result for SDG 12 Responsible consumption and production.
In the social dimension, the SIDS mostly come out as could be expected by developing countries. While results for education are relatively positive, substantial work remains to be done regarding health and two challenges stand out. The most striking finding is the extent to which the SIDS group as a whole performs badly in terms of adult obesity. This is generally the case for 23 SIDS across all three ocean groups when compared to the GA and the result holds for 11 SIDS even when we compare to the HICs. Further, a very clear and negative result is the poor performance generally when it comes to gender equality, especially when measured using the share of women in national parliaments as the indicator. Achieving better gender equality is something that the SIDS can address independently of their progress with other development ambitions.
Finally, the SIDS overall come out reasonably well in the environmental dimension which is not surprising considering that poor countries generally leave a smaller ecological footprints than richer ones, and especially considering their low levels of industrialisation. The most negative outcome is for SDG 15 Life on land which can be explained by their limited landmass. Meanwhile, the SIDS score high in ocean biodiversity and health, but ocean fish stocks seem to be mostly concentrated on the Pacific rather than the Caribbean, where the distance to the continent is much lower. Focusing on ocean territories and marine resources exposes the SIDS' potentially greatest future opportunities -to develop a sustainable Blue Economy-led pathway towards socio-economic development.
Based on our broad mapping of SDG performance, we suggest four concrete policy recommendations. First, it is urgent that SIDS further develop their statistical capacity to enable more in-depth understanding of their individual development performance and potential for collaborations. Second, several SIDS, especially in the Pacific, urgently need to address adult obesity to improve the health status of their populations. Third, the work of promoting gender equality needs to be lifted as a prioritised area. Finally, all SIDS have a window of opportunity to push their development of Blue Economy activities.