Navigating the 2030 Agenda: A Holistic Evaluation of Global Progress and Systemic Barriers Across All 17 Sustainable Development Goals

Author: Isabella Rossi

Affiliation: Swiss International University (SIU)

ORCID ID: 0009-0009-1291-9122

Submitted 18 April 2026; Revised 14 June 2026; Accepted 21 July 2026; Available online 08 August 2026; Version of Record 08 August 2026.

Doi: https://doi.org/10.65326/u7y.SpecSDG10010

Volume 3, December 2026, (SpecSDG10010)

SpecSDG10010.pdf

Download PDF • 657KB

Abstract

With five years remaining before the 2030 deadline, the United Nations reports that 35% of assessable Sustainable Development Goal targets are on track or advancing moderately and that 18% have regressed below their 2015 baseline. Explanations for the shortfall divide into three diagnoses that are rarely tested against one another: an investment gap now estimated near four trillion dollars a year, a governance architecture that sets goals without binding anyone to them, and a measurement system that cannot see roughly half of what it is meant to track. This policy analysis appraises the progress record against the monitoring infrastructure that produces it, then examines the four constraints that the literature treats as parallel obstacles: commitment design, measurement and visibility, selective national implementation, and financing. The analysis finds the constraints are causally coupled rather than additive. Indicator selection removed the distributional and environmental content of several goals; the goals that lost measurability were subsequently deprioritised in national reviews; weak accountability then permitted the shortfall to be attributed to insufficient finance; and financing responses concentrated on the goals that remained visible. Shocks act on this cycle as an amplifier rather than an independent cause. The paper sets out three feedback loops that sustain the cycle, identifies the points at which each can be cut, and argues that a post-2030 architecture that fixes financing without fixing bindingness and measurement will reproduce the same shortfall.

Keywords: 2030 agenda, sustainable development goals, policy coherence, sdg indicators, development finance, global governance.

1.     Introduction

The 2030 Agenda is closer to its deadline than to its adoption, and the arithmetic has stopped being encouraging. The United Nations (2025) reports that 35% of assessable targets are on track or making moderate progress, nearly half are advancing too slowly, and 18% now sit below where they stood in 2015. Xing et al. (2025), assessing 117 indicators across 167 countries, project a global SDG score near 63% by 2030 and calculate that meeting the targets would require roughly 4% annual improvement, several times the observed rate. Neither of those figures describes a programme that is behind schedule. They describe one that will not arrive.

Explanations for the shortfall fall into three groups that seldom confront one another. The financing diagnosis is the most institutionally established: the United Nations Conference on Trade and Development (UNCTAD, 2023) puts the annual investment gap in developing countries at roughly four trillion dollars, up from two and a half trillion when the goals were adopted, and the Organisation for Economic Co-operation and Development (OECD, 2025) reviews the Addis Ababa Action Agenda across seven action areas as the principal financing response. A governance diagnosis locates the problem elsewhere. Biermann et al. (2017) described the goals from the outset as governance through goal-setting, inclusive in formation but non-binding and institutionally thin, and Biermann et al. (2022), meta-analysing more than 3,000 studies published between 2016 and April 2021, concluded that the political impact of the goals has been largely discursive. A third diagnosis concerns measurement: Dang and Serajuddin (2020) find data available for just over half of the indicators, and for only 19% of what tracking progress comprehensively across countries and over time would require.

Each diagnosis is supported. None has displaced the others, and the reason is that they are usually studied apart. A financing analysis takes the goals and their monitoring as given and asks what resources would close the gap. A governance analysis takes the shortfall as given and asks what institutional design produced it. A measurement analysis asks what the numbers can support and stops there. The result is a catalogue of parallel obstacles and a sequence of single-constraint remedies, with the shortfall re-attributed each time to whichever constraint the analyst happens to study.

I argue that these are not parallel obstacles. They are coupled, and the coupling runs in a direction that makes each one harder to fix than it looks in isolation. Indicator selection stripped distributional and environmental content out of several goals during the negotiation that produced them (Fukuda-Parr, 2019; Elder & Olsen, 2019). Goals that lost measurability lost visibility in national reporting, and national reviews then concentrated on the goals that were easiest to show progress on (Forestier & Kim, 2020). Weak bindingness meant that concentration carried no cost, which left the shortfall available for attribution to insufficient finance. Financing responses in turn flow toward what can be counted and reported. The cycle closes, and each rotation makes the next one more likely.

Monsod et al. (2023) supply the sharpest reason to take this seriously. Estimating endogenous debt limits for climate-vulnerable developing countries, they find fiscal space in fact fairly ample for many of them, and argue that conflating standard debt-sustainability thresholds with genuine solvency limits traps otherwise solvent governments. If that assessment holds even partially, the financing gap is more than a resource constraint. It is in part a governance and analytical construct, produced by the same institutional weaknesses it is invoked to explain.

The paper does three things. It appraises the reported progress record against the measurement infrastructure that generates it, so that what is known is separated from what is merely unreported. It analyses the four constraints in turn and specifies how each transmits to the next. And it sets out the resulting cycle as a diagnostic, identifying where the loops can be cut and what that implies for the years remaining and for whatever architecture follows 2030.

2.     Progress at the Two-Thirds Mark and What the Record Can Support

Any evaluation of the 2030 Agenda inherits the limits of the system that measures it, so the record and its basis have to be read together. Table 1 sets out the principal assessments alongside their coverage, metric and caveats.

The headline assessments agree on direction while differing on magnitude. The United Nations (2025) gives the official stocktake: 35% on track or moderate, nearly half too slow, 18% in reverse. Xing et al. (2025) add a finding that the aggregate obscures, namely that progress depends on where a country starts. Among low-scoring indicators, 25% of countries advanced against 16% that regressed; among high-scoring indicators the figures were 16% and 10%. Improvement is happening, and it is happening unevenly enough that the global average describes almost no one.

Projection work reached compatible conclusions well before the reversals of the 2020s. Moyer and Hedden (2020) modelled nine indicators across six human development goals under a middle-of-the-road scenario and found that 43% of 1,674 country-indicator pairs had already met their target values in 2015, with only 53% projected to do so by 2030. Sanitation, upper secondary completion and child underweight were furthest off track, and 28 countries were identified as most vulnerable. A five-year gain of ten percentage points on goals that were already partly achieved at adoption is not the trajectory the Agenda was designed around, and that projection assumed no pandemic.

The measurement basis is weaker than the confidence of the headline numbers suggests. MacFeely (2020) examined the framework of 232 indicators adopted by the United Nations Statistical Commission in March 2017 and set out both the compilation burden it imposes and the unanticipated consequences of the measurement mechanism itself. Dang and Serajuddin (2020) quantified the shortfall: data exist for just over half the indicators, and for 19% of what comprehensive tracking would need, while ambiguous indicator wording lets different evaluation methods reach different verdicts on the same country. Beegle et al. (2025) show that absent data is only part of the problem. Countries reported on average 38% of the 50 SDG gender indicators in at least one year between 2016 and 2020, and sex-disaggregating the 32 indicators for which population estimates already exist would raise coverage to 47% without collecting anything new. Poorer countries reported no worse than high-income ones, which locates the failure in reporting practice rather than in statistical poverty alone.

Composite indices partly compensate and partly conceal. Kynčlová et al. (2020) built an SDG-9 index for 128 economies over 2000 to 2016 from the official framework and used it to expose the specific dimensions on which countries lag, which is what a well-constructed index should do. The same construction also produces a single number that travels further than its components, and a single number is what a ministry reports.

Two conclusions follow for the rest of the analysis. Assessments of the Agenda's progress are more reliable for the goals that were easiest to measure at the outset, which are broadly the socioeconomic ones, and least reliable for the distributional, environmental and institutional goals. And the “18% regressed” figure is a floor rather than an estimate, because regression on an unreported indicator does not register at all. The Independent Group of Scientists appointed by the Secretary-General (2023) concluded that incremental and fragmented change will not deliver the goals in the time remaining. The measurement record suggests the situation is somewhat worse than the incrementalism visible in the data.

Table 1

Principal assessments of 2030 Agenda progress and the measurement basis on which they rest

Assessment	Coverage	Metric	Headline finding	Measurement caveat
United Nations (2025)	All assessable targets, global	Share of targets on track, slow, or regressing	35% on track or moderate; nearly half too slow; 18% below the 2015 baseline	Assessable targets only; regression on unreported indicators does not register
Xing et al. (2025)	117 indicators, 167 countries	Country-level advance or regression by baseline score	Low-scoring indicators: 25% advanced, 16% regressed; high-scoring: 16% and 10%; global score projected near 63% by 2030	Projection assumes continuation of observed rates
Moyer and Hedden (2020)	9 indicators, 6 human development goals, 1,674 country-indicator pairs	Share of pairs meeting target values under an SSP2 baseline	43% met targets in 2015, 53% projected by 2030; 28 countries most vulnerable	Pre-pandemic scenario; human development goals only
Kynčlová et al. (2020)	128 economies, 2000–2016	Composite index of inclusive and sustainable industrial development	Industrialised economies lead; index isolates the dimensions on which countries lag	Single goal (SDG 9); composite scores travel further than their components
Dang and Serajuddin (2020)	Full official indicator set and UN SDG database	Data availability	Data exist for just over half of indicators, and for 19% of what comprehensive tracking would require	Ambiguous indicator wording lets methods differ on the same country
Beegle et al. (2025)	50 SDG gender indicators, 2016–2020	National reporting coverage	Countries reported 38% on average; disaggregating 32 existing indicators would raise coverage to 47%	Reporting failure, not data absence, drives much of the gap
MacFeely (2020)	232-indicator framework adopted March 2017	Compilation burden and measurement design	Documents the statistical challenge and the unanticipated consequences of the measurement mechanism	Analysis of the framework rather than of country performance
Independent Group of Scientists (2023)	Agenda-wide science assessment	Qualitative appraisal of transformation	Incremental and fragmented change will not deliver the goals by 2030	Assessment rather than measurement; no indicator-level estimates

Note. Compiled by the author from the cited assessments. Percentages are reproduced as reported by each source and are not comparable across rows, since coverage, metric and reference period differ. The caveat column records the limitation each source states or that follows directly from its design.

3.     Analytical Framework: Four Coupled Constraints

The analysis that follows appraises four constraints against four criteria, and the criteria are stated here so that the judgements can be checked.

Constraints were selected on the basis that each is independently documented in the peer-reviewed or intergovernmental literature as an obstacle to SDG achievement, and that each operates on the Agenda as a whole rather than on a single goal. That yields four: commitment design, meaning the bindingness and institutional backing of the goals; measurement and visibility, meaning what the indicator framework can see; selective implementation, meaning which goals national governments actually pursue; and financing and fiscal space, meaning the resources and the fiscal rules that govern their availability. Shocks are treated not as a fifth constraint but as an amplifier acting on all four, following the polycrisis account of Lawrence et al. (2024).

Each constraint is appraised on four criteria: the strength of the evidence that it binds, drawn only from sources that state their data and method; the mechanism by which it blocks progress, stated explicitly rather than asserted; its transmission to the other three constraints, which is the criterion that distinguishes this analysis from a catalogue; and the tractability of relief, meaning whether relief requires new resources, new rules, or only new practice.

Sources were identified through scholarly database searching combining ("Sustainable Development Goals" OR "2030 Agenda") with terms for progress, implementation, governance, indicators, financing, Voluntary National Reviews and policy coherence, supplemented by direct retrieval of intergovernmental assessments that indexed databases cover poorly. The window ran from 2015 to 2026. Inclusion required a peer-reviewed article or an official intergovernmental assessment, an explicit empirical or documentary basis, an Agenda-wide rather than single-sector claim, and publication in English. Every reference was verified against CrossRef metadata or the issuing organisation's own page before use, with authors, year, outlet and pagination confirmed; no claim rests on a source whose metadata could not be confirmed.

The design is a policy analysis of documented institutions, monitoring arrangements and programmes rather than an empirical study, and it generates no new data. Its central proposition — that the four constraints form a reinforcing cycle — is a causal claim assembled from evidence collected for other purposes, and it is stated as an interpretation open to disconfirmation rather than as a measured result. Two safeguards apply. Evidence that cuts against the argument is engaged where it exists, most directly Monsod et al. (2023) against the financing consensus and Glass and Newig (2019) against the assumption that governance quality broadly explains achievement. And the strength of each link in the cycle is stated separately, so that a reader who rejects one link can see what survives.

4.     Commitment Design: Goals Without Obligation

The Agenda's institutional form was a deliberate choice, and its consequences are now measurable. Biermann et al. (2017) characterised it as governance through goal-setting: an inclusive formation process, non-binding goals, weak institutional arrangements, and extensive national leeway, with success dependent on states formalising commitments, strengthening global arrangements, translating goals into national contexts and integrating sectoral policy. Almost none of those conditions has been met at scale.

The evidence on what the form produced is now substantial. Biermann et al. (2022) found across more than 3,000 studies that the political impact of the goals has been largely discursive, altering how actors talk about sustainable development while legislative change and reallocated resources remain rare. Hickmann et al. (2024), a scoping review by 32 governance scholars, identified five arenas of observable effect — global governance, national policy integration, subnational initiatives, private governance and education — and concluded that effects remain patchy and often symbolic. Two independent reviews of a large literature converging on the same verdict is about as strong as evidence about political impact becomes.

Glass and Newig (2019) qualify this usefully. Testing which governance attributes explain SDG achievement across 41 high and upper-middle income countries, they found that of participation, policy coherence, reflexivity, adaptation and democratic institutions, only democratic institutions and participation contributed alongside economic power, education and geography. Policy coherence, the attribute that SDG governance discourse has emphasised most heavily, did not register. That finding should temper any expectation that coherence machinery alone will move outcomes, and it points attention toward the political conditions under which coherence is actually enforced.

It is worth being precise about what the design achieved, because dismissing it would misstate the counterfactual. The inclusive formation process that Biermann et al. (2017) describe is what produced a universal agenda covering environment, inequality and institutions rather than a donor-defined poverty list, and a binding treaty covering that scope would not have been adopted. The trade the drafters made was breadth of ambition for enforceability. What the subsequent evidence shows is that the cost of that trade was underestimated: the goals inherited the political reach of a treaty and the compliance apparatus of a communiqué, and states have used the resulting leeway more than they have used the ambition.

The transmission from this constraint to the others is straightforward. Where nothing is binding, reporting substitutes for compliance, and what is reported is what can be measured. Commitment design therefore hands the next constraint its authority: in a non-binding regime the indicator framework does not merely describe performance, it defines it.

5.     Measurement and Visibility: What the Indicator Framework Removed

The indicator framework is usually treated as a technical instrument that lags political ambition. The evidence suggests it also shaped that ambition, and did so during the negotiation rather than afterwards.

Fukuda-Parr and McNeill (2019) document systematic slippage between the goals produced by the Open Working Group and the targets and indicators eventually selected, and argue that indicator selection, presented throughout as technical, is a political act that reorients what a goal means. Fukuda-Parr (2019) traces the mechanism for SDG 10. A goal that appears to set a strong norm on reducing inequality was rendered into targets and indicators built around shared prosperity rather than distributional measures such as the Gini coefficient or the Palma ratio, converting a commitment to reduce extreme inequality into a commitment to inclusion. Elder and Olsen (2019) find the parallel pattern on the environmental side: environmental concerns were incorporated broadly but many environment-related targets were housed inside non-environmental goals, planetary boundaries and beyond-GDP framings were rejected in favour of decoupling and resource efficiency, and environmental elements of many targets were omitted or weakened at the indicator stage.

Bexell (2024) shows the same process operating after adoption rather than during it. Tracing what she terms indicator reporting trajectories through the case of SDG 17 multistakeholder partnerships, she finds that globally agreed commitments can disappear quietly across long follow-up cycles for want of data and measurability, producing policy shrinking in place of accountability. A commitment that cannot be reported is not formally abandoned. It simply stops appearing.

MacFeely (2020) adds the administrative dimension that makes the pattern self-sustaining. Compiling 232 indicators is a burden that falls on national statistical offices whose capacity was built for a much smaller monitoring task, and the measurement mechanism itself produces consequences its designers did not anticipate. Offices under that load rationalise toward the indicators that are already collected for other purposes, which are the economic and social series that predate the Agenda. The environmental and institutional indicators, which would require new instruments and new administrative relationships, are the ones deferred. No decision to neglect them is ever taken; the neglect is the aggregate of reasonable triage decisions made under a workload nobody costed at adoption.

Set beside the coverage failures documented earlier, this yields the constraint's mechanism. The indicator framework does not fail uniformly. It fails hardest on distribution, environment and institutions, which are the goals whose achievement would require redistributive or regulatory action, and it fails least on the socioeconomic indicators that improve with growth. The framework thereby encodes a bias toward the goals that are politically cheapest to pursue, and in a non-binding regime that bias becomes a de facto prioritisation. Whether the bias was intended is not something the sources establish, and the analysis here does not claim it. What they do establish is that it operates.

6.     Selective Implementation and the Erosion of Indivisibility

Indivisibility is the Agenda's foundational claim about itself, and national practice has not honoured it. Forestier and Kim (2020) content-analysed the Voluntary National Reviews of 19 countries across income levels and found SDG 1 and SDG 8 by far the most widely prioritised, describing the pattern as cherry-picking that defies the integrated character of the Agenda. Those two goals are also among the best measured and the most compatible with existing growth strategies, which is what the preceding section predicts.

National studies show the selection is not arbitrary but domestically driven. Horn and Grugel (2018), interviewing national and Quito city policymakers in Ecuador, found engagement concentrated on inequality and inclusive-cities targets and refracted through the country's own Buen Vivir agenda, concluding that states do not adopt the SDGs as a template so much as filter them through decentralised structures and prior political preferences. Jönsson and Bexell (2021) find in Tanzania that government, civil society and parliamentary engagement drive localisation while unclear allocation of responsibility, weak coordination and shrinking democratic space obstruct it, and that localisation demands awareness well beyond elite circles. Koff and Häbel (2022), analysing normative coherence across world regions, conclude that regions matter as intermediaries between global and local levels but that normative coherence remains unachieved, because overriding political priorities and narrowly technical readings of coherence crowd out the normative content of the Agenda.

The methods that would expose the cost of selectivity exist and are not being used. Allen et al. (2021) reviewed more than 150 publications to identify 22 science-based approaches recommended for national implementation, then examined 56 Voluntary National Reviews and found cross-fertilisation mainly in monitoring and evaluation, with methods for assessing interlinkages and national transformations largely absent. Pradhan et al. (2024) make the same point from the science side, arguing that the integrated character of the goals is being undermined by siloed implementation and identifying SDG interactions, modelling and decision-support tools as the science-policy foci that would operationalise integration. Sachs et al. (2019) offer the structural alternative, organising the Agenda into six transformations that fit existing government machinery while preserving interdependence. Lusseau and Mancini (2019) supply the reason selectivity carries a real cost rather than merely an aesthetic one: the contribution of individual goals to overall progress varies systematically with country income, and poverty and inequality reduction generate compound positive effects, so choosing the wrong subset forgoes gains elsewhere.

Selectivity transmits to financing directly. A government that has narrowed its effective agenda to the goals it can show progress on will cost that agenda, not the full one, and the difference between the two appears in aggregate accounts as an unmet financing need.

7.     Financing and Fiscal Space: A Gap That Is Partly Constructed

The financing diagnosis is the one with institutional weight behind it, and the numbers are large. UNCTAD (2023) reports the annual SDG investment gap in developing countries widening from 2.5 trillion dollars in 2015 to roughly 4 trillion, with the deepest shortfalls in energy, water and transport infrastructure. Gaspar et al. (2019) estimate the additional annual public spending on health, education, roads, electricity and water and sanitation that countries would need by 2030 and analyse the financing strategies open to low-income and emerging economies. OECD (2025) reviews progress across the seven action areas of the Addis Ababa Action Agenda as an input to the fourth financing-for-development conference.

The gap is real. What it means is contested. Monsod et al. (2023) apply an endogenous debt-limit measure to climate-vulnerable developing countries and find fiscal space fairly ample for many of them, arguing that treating standard debt-sustainability thresholds as limits to fiscal space leaves solvent governments unable to borrow for adaptation and development. On that reading, part of the gap is produced by the analytical conventions used to assess it, which is a governance failure wearing the costume of a resource shortage.

The composition of existing spending points the same way. Naidoo and Fisher (2020) note roughly 4.7 trillion dollars in annual fossil-fuel subsidies and argue for redirecting that flow along with part of military spending, alongside prioritising goal combinations that deliver several objectives at once. Barbier and Burgess (2020) reach a compatible conclusion from the post-pandemic fiscal position, arguing that with less financing available the priority should be low-cost measures that advance several goals simultaneously, including redirecting fossil fuel and irrigation subsidies toward energy and water poverty and a tropical carbon tax to fund natural climate solutions. Both arguments assume something the preceding sections have shown to be weak: an institutional capacity to reallocate against entrenched interests, and an ability to identify which goal combinations pay off jointly. The first requires bindingness. The second requires the interlinkage methods that Allen et al. (2021) found missing from national reviews.

The institutional response has been organised around resource mobilisation rather than around the conditions under which resources translate into outcomes. The seven action areas reviewed by OECD (2025) — domestic public resources, private business and finance, development co-operation, trade, science and technology, debt, and systemic issues — describe an architecture built to increase and channel flows. That is the correct architecture for a shortage of money. It contains no instrument that binds a recipient government to the goals it deprioritised, and no mechanism that improves the reporting on which allocation decisions depend. A financing conference is therefore capable of closing a financing gap and incapable of closing the gap between what states pledged and what they pursue, which are not the same shortfall even though the same number is used to describe both.

The financing constraint is therefore the least independent of the four. Its magnitude depends on which agenda is being costed, which depends on selectivity; its distribution depends on debt rules that are analytical conventions rather than physical limits; and its deployment depends on measurement, because money moves toward what can be reported.

8.     Shocks as Amplifier

Treating the pandemic and its successors as the explanation for the shortfall would misread the sequence, since the projection literature was already pessimistic before 2020. What shocks do is convert slow constraints into visible reversals.

Naidoo and Fisher (2020) estimated early in the pandemic that roughly two-thirds of the 169 targets were under threat. Shulla et al. (2021), working from moderated expert focus groups, identified a pandemic-driven pattern of interconnection centred on health, education, decent work, consumption and climate action, with spillovers threatening gender equality, infrastructure, inequality reduction, sustainable cities and partnerships. Lawrence et al. (2024) give the general form, defining global crisis as fast-moving triggers combining with slow-moving stresses to push a global system out of equilibrium, and identifying common stresses, domino effects and inter-systemic feedbacks as the pathways of entanglement.

Read against the four constraints, the amplification is specific rather than diffuse. Shocks raise financing needs while contracting fiscal space, which widens the measured gap without any change in the underlying agenda. They push governments toward emergency prioritisation, which deepens selectivity. They interrupt data collection, which thins the evidence base exactly when reallocation decisions are being made. A system with binding obligations, complete measurement and integrated planning would absorb such a shock. The Agenda as constituted transmits it.

9.     The Barrier Reinforcement Cycle

The four constraints and the shock amplifier assemble into the cycle shown in Figure 1, and the value of stating it as a cycle rather than a list is that it explains why single-constraint remedies have underperformed. Table 2 summarises each constraint with its evidence, blocking mechanism and the point at which it can be relieved.

Figure 1

The Barrier Reinforcement Cycle: four coupled constraints, the three loops that sustain them, and shocks as amplifier

Note. The four constraints and the evidence supporting each are set out in Table 2. The three loops are stated at different evidential strengths in the text: the visibility loop rests on the firmest evidence, the attribution loop is the most interpretive.

Three feedback loops carry the cycle. The visibility loop runs from indicator design to national prioritisation and back: goals whose content was weakened at the indicator stage attract less national reporting, thin reporting further weakens the case for strengthening those indicators, and the goals recede from view (Fukuda-Parr, 2019; Bexell, 2024; Forestier & Kim, 2020). The attribution loop runs from weak bindingness to financing: because no actor is accountable for a missed target, the shortfall is available for attribution to insufficient finance, which is the explanation that assigns responsibility to no one in particular and requires no institutional change (Biermann et al., 2022; Hickmann et al., 2024; UNCTAD, 2023). The allocation loop runs from financing back to visibility: resources flow toward outcomes that can be counted and reported, which improves performance and data on the already-visible goals and widens the gap against the rest (Beegle et al., 2025; Dang & Serajuddin, 2020).

Each loop is a claim of different strength, and they should not be accepted together. The visibility loop rests on the firmest evidence, since indicator weakening, reporting gaps and national prioritisation patterns are each separately documented and their alignment is close. The attribution loop is the most interpretive: the sources establish that impact has been discursive and that the financing framing dominates, but no study demonstrates that the second follows from the first. The allocation loop sits between the two, supported by the concentration of both data coverage and national attention on the same subset of goals but not by direct evidence on where money went.

The cycle also identifies where it can be cut, which is the practical payoff. Cutting at measurement is the cheapest available intervention and the most immediately achievable: Beegle et al. (2025) show that sex-disaggregating existing data would lift gender indicator coverage from 38% to 47% without new collection, and comparable reporting gains are likely elsewhere. Cutting at selectivity requires no new resources either, only the use of interlinkage methods that already exist and that Allen et al. (2021) found absent from national reviews. Cutting at commitment design would do the most and is the least likely, since it would require what Biermann et al. (2017) called for at the outset and states have consistently declined. Cutting at financing alone, which is where institutional effort has concentrated, is the intervention the cycle predicts will underdeliver, because money entering a system with weak bindingness, incomplete measurement and selective implementation will follow the same channels that produced the imbalance.

Table 2

The four constraints: evidence, blocking mechanism, and where the cycle can be cut

Constraint	Principal evidence	How it blocks progress	Where it can be cut
Commitment design	Biermann et al. (2017, 2022); Hickmann et al. (2024); Glass and Newig (2019)	Non-binding goals with weak institutional backing make reporting a substitute for compliance, so missed targets carry no cost	Formalised commitments and enforcement; highest effect, lowest political feasibility
Measurement and visibility	Fukuda-Parr (2019); Fukuda-Parr and McNeill (2019); Elder and Olsen (2019); MacFeely (2020); Dang and Serajuddin (2020); Beegle et al. (2025); Bexell (2024)	Distributional, environmental and institutional content was weakened at indicator selection and is under-reported afterwards, so the hardest goals become invisible	Report existing data and disaggregate what is already collected; cheapest available intervention
Selective implementation	Forestier and Kim (2020); Horn and Grugel (2018); Jönsson and Bexell (2021); Koff and Häbel (2022); Allen et al. (2021); Pradhan et al. (2024)	National reviews concentrate on the best-measured, growth-compatible goals, so indivisibility erodes without any formal decision	Require reviews to state deprioritised goals; apply existing interlinkage methods
Financing and fiscal space	UNCTAD (2023); Gaspar et al. (2019); OECD (2025); Monsod et al. (2023); Naidoo and Fisher (2020); Barbier and Burgess (2020)	Resource shortfalls are real, but the estimated gap depends on which agenda is costed and on debt conventions that function as rules rather than limits	Revise solvency assessment and redirect existing subsidy flows; ineffective if pursued alone
Shocks (amplifier)	Naidoo and Fisher (2020); Shulla et al. (2021); Lawrence et al. (2024)	Raise financing needs while contracting fiscal space, deepen emergency prioritisation, and interrupt data collection	Not a target for intervention; the case for relieving the other four before the next shock

Note. Compiled by the author. Shocks are listed for completeness as an amplifier acting on the four constraints rather than as a fifth constraint. The final column states the intervention implied by the analysis, not a costed policy proposal.

10.  Discussion

What the cycle implies for the remaining period

Five years is too short to redesign the Agenda's institutional form, which makes the low-cost cuts the ones worth pressing. Reporting existing data that is currently unreported, sex-disaggregating indicators where the population estimates already exist, and requiring Voluntary National Reviews to state which goals were deprioritised and why would together change what is visible without changing what is owed. Visibility is not accountability, but in a non-binding regime it is the closest available substitute, and the evidence in this analysis suggests it is the input on which the other constraints depend.

The finding also reframes the financing conversation rather than dismissing it. The investment gap is genuine and the deficits in energy, water and transport infrastructure are physical. The claim here is narrower: that the gap's magnitude is partly endogenous to which agenda is costed and to which debt conventions are applied, and that closing it without addressing the other three constraints will fund the goals that were already advancing. Monsod et al. (2023) show that at least one component of the perceived shortage is a rule rather than a resource, and rules can be revised on a shorter timescale than tax bases can be grown.

Implications for a post-2030 architecture

Design work on what follows 2030 is underway, and the cycle carries three implications for it. A successor framework that retains non-binding goals while improving indicators will strengthen the measurement of an unenforced commitment. One that binds without fixing measurement will bind states to the goals that were already easiest to show. One that fixes both while leaving the unit of coordination at 17 separately reported goals will reproduce selectivity, because selection pressure operates on whatever the reporting unit is. The six transformations of Sachs et al. (2019) and the interaction-based approach of Pradhan et al. (2024) both suggest organising commitment and reporting around delivery systems rather than goal numbers, which would make partial implementation visible as an incomplete transformation rather than as progress on a subset.

Legitimacy belongs in this design conversation too. Sénit et al. (2017) assessed three civil society consultations in which the United Nations consulted nearly ten million people and found their democratising potential limited by sociodemographic bias and under-representation of developing-country actors, with clear objectives, adequate participant time and formal linkage to negotiations mattering more than resources or open access. Sénit (2020) shows that civil society influence on outcomes concentrated in informal and exclusive venues early in the process, where the groups with access were least representative. A commitment to leave no one behind that was drafted through venues structured this way starts with a representational deficit that no indicator can correct later.

What would falsify the cycle

An interpretive construct earns its place by being refusable, so the disconfirming evidence should be named. The cycle predicts that national attention tracks measurability; a finding that Voluntary National Reviews give sustained priority to goals with poor indicator coverage would break the visibility loop. It predicts that financing follows reportability; disbursement data concentrated on poorly measured goals would break the allocation loop. It predicts that attributions of failure shift toward resources as accountability weakens; a documentary record in which institutional explanations hold steady or grow would break the attribution loop. The cycle also predicts that improving reporting raises subsequent attention, so a country that closed a coverage gap without any change in national priority would count against it. None of these tests has been run, and the argument here should be held only as firmly as that admission allows.

Limitations

This is a policy analysis of documented institutions and published assessments, not an empirical study, and its central construct is an interpretation. The cycle is assembled from evidence gathered for other purposes, and while each constraint is separately documented, the causal links between them are inferred from alignment rather than demonstrated by design. The three loops are stated at different strengths for that reason, and the attribution loop in particular could be wrong without the rest of the analysis failing. The evidence base is uneven in a way that matters: governance and indicator scholarship is concentrated in a small number of European and North American research groups, national implementation evidence rests on a handful of country cases, and English-language publication bias excludes work that would likely complicate the selectivity findings. The progress figures reproduced here inherit the coverage limits the analysis itself identifies, which means the assessment of how far off track the Agenda is cannot be more precise than the data that would establish it. No quantification of the cycle is attempted, and none of the proposed interventions is costed.

Future research

Three lines would test what this analysis infers. The attribution loop is checkable: coding how national and multilateral documents explain missed targets over time, and testing whether financing explanations displace institutional ones as bindingness weakens, would either support the loop or break it. The allocation loop needs disbursement data mapped against indicator coverage, to establish whether resources do concentrate on well-measured goals. And the measurement cut proposed here has a natural experiment available, since countries differ in how much existing data they report; comparing subsequent national attention to goals whose coverage improved against those where it did not would show whether visibility moves priority in the direction the cycle assumes. Beyond testing, the more useful work is design: specifying what a binding, transformation-organised, fully reported successor framework would look like in enough operational detail that states are refusing something concrete rather than an abstraction.

11.  Conclusion

The 2030 Agenda will miss most of its targets, and the reasons are better understood separately than together. Money is short, the goals bind no one, and the measurement system cannot see roughly half of what it was built to track. Treating these as three problems has produced three sets of remedies, of which only the financing one has attracted sustained institutional effort, and it is the one the evidence suggests is least independent of the others.

The account offered here is that the constraints hold each other in place. Indicators that were weakened during negotiation made certain goals hard to see; invisibility permitted national selectivity; selectivity went uncorrected because nothing binds; and the resulting shortfall was attributed to a financing gap whose magnitude depends on which agenda is costed and which debt conventions are applied. Shocks did not create this arrangement, they revealed how little it absorbs. What follows from the analysis is a narrow practical claim and a broader design one. Reporting the data that already exists is the cheapest available intervention and the one on which the others depend. And any successor to the 2030 Agenda that improves financing while leaving bindingness and measurement as they are will be a better-funded version of the same shortfall.

Declarations

Funding. This research received no external funding.

Conflicts of interest. The author declares no conflict of interest.

Ethics statement. This study analyses published literature, official statistics and intergovernmental policy documents. It involved no human participants, no animal subjects and no primary data collection, and therefore required no ethical approval.

Data availability. No new data were generated. All sources analysed are published works listed in the reference list and identified by digital object identifier or public repository address.

References

• Allen, C., Metternicht, G., & Wiedmann, T. (2021). Priorities for science to support national implementation of the sustainable development goals: A review of progress and gaps. Sustainable Development, 29(4), 635-652. https://doi.org/10.1002/sd.2164

• Barbier, E. B., & Burgess, J. C. (2020). Sustainability and development after COVID-19. World Development, 135, Article 105082. https://doi.org/10.1016/j.worlddev.2020.105082

• Beegle, K., Serajuddin, U., Stacy, B., & Wadhwa, D. (2025). Missing SDG gender indicators. World Development, 196, Article 106957. https://doi.org/10.1016/j.worlddev.2025.106957

• Bexell, M. (2024). Indicator accountability or policy shrinking? Multistakeholder partnerships in reviews of the sustainable development goals. Global Policy, 15(2), 276-287. https://doi.org/10.1111/1758-5899.13348

• Biermann, F., Hickmann, T., Sénit, C.-A., Beisheim, M., Bernstein, S., Chasek, P., Grob, L., Kim, R. E., Kotzé, L. J., Nilsson, M., Ordóñez Llanos, A., Okereke, C., Pradhan, P., Raven, R., Sun, Y., Vijge, M. J., van Vuuren, D., & Wicke, B. (2022). Scientific evidence on the political impact of the Sustainable Development Goals. Nature Sustainability, 5(9), 795-800. https://doi.org/10.1038/s41893-022-00909-5

• Biermann, F., Kanie, N., & Kim, R. E. (2017). Global governance by goal-setting: The novel approach of the UN Sustainable Development Goals. Current Opinion in Environmental Sustainability, 26-27, 26-31. https://doi.org/10.1016/j.cosust.2017.01.010

• Dang, H.-A. H., & Serajuddin, U. (2020). Tracking the sustainable development goals: Emerging measurement challenges and further reflections. World Development, 127, Article 104570. https://doi.org/10.1016/j.worlddev.2019.05.024

• Elder, M., & Olsen, S. H. (2019). The design of environmental priorities in the SDGs. Global Policy, 10(S1), 70-82. https://doi.org/10.1111/1758-5899.12596

• Forestier, O., & Kim, R. E. (2020). Cherry-picking the Sustainable Development Goals: Goal prioritization by national governments and implications for global governance. Sustainable Development, 28(5), 1269-1278. https://doi.org/10.1002/sd.2082

• Fukuda-Parr, S. (2019). Keeping out extreme inequality from the SDG agenda - The politics of indicators. Global Policy, 10(S1), 61-69. https://doi.org/10.1111/1758-5899.12602

• Fukuda-Parr, S., & McNeill, D. (2019). Knowledge and politics in setting and measuring the SDGs: Introduction to special issue. Global Policy, 10(S1), 5-15. https://doi.org/10.1111/1758-5899.12604

• Gaspar, V., Garcia-Escribano, M., Prady, D., & Soto, M. (2019). Fiscal policy and development: Human, social, and physical investments for the SDGs (Staff Discussion Note No. 2019/003). International Monetary Fund. https://www.imf.org/en/publications/staff-discussion-notes/issues/2019/01/18/fiscal-policy-and-development-human-social-and-physical-investments-for-the-sdgs-46444

• Glass, L.-M., & Newig, J. (2019). Governance for achieving the Sustainable Development Goals: How important are participation, policy coherence, reflexivity, adaptation and democratic institutions? Earth System Governance, 2, Article 100031. https://doi.org/10.1016/j.esg.2019.100031

• Hickmann, T., Biermann, F., Sénit, C.-A., Sun, Y., Bexell, M., Bolton, M., Bornemann, B., Censoro, J., Charles, A., Coy, D., Dahlmann, F., Elder, M., Fritzsche, F., Gehre Galvão, T., Grainger-Brown, J., Inoue, C., Jönsson, K., Koloffon Rosas, M., Krellenberg, K., ... Weiland, S. (2024). Scoping article: Research frontiers on the governance of the Sustainable Development Goals. Global Sustainability, 7, Article e7. https://doi.org/10.1017/sus.2024.4

• Horn, P., & Grugel, J. (2018). The SDGs in middle-income countries: Setting or serving domestic development agendas? Evidence from Ecuador. World Development, 109, 73-84. https://doi.org/10.1016/j.worlddev.2018.04.005

• Independent Group of Scientists appointed by the Secretary-General. (2023). Global Sustainable Development Report 2023: Times of crisis, times of change - Science for accelerating transformations to sustainable development. United Nations. https://sdgs.un.org/gsdr/gsdr2023

• Jönsson, K., & Bexell, M. (2021). Localizing the Sustainable Development Goals: The case of Tanzania. Development Policy Review, 39(2), 181-196. https://doi.org/10.1111/dpr.12497

• Koff, H., & Häbel, S. (2022). Normative coherence for development: What relevance for responsive regionalism? Development Policy Review, 40(S1), Article e12586. https://doi.org/10.1111/dpr.12586

• Kynčlová, P., Upadhyaya, S., & Nice, T. (2020). Composite index as a measure on achieving Sustainable Development Goal 9 (SDG-9) industry-related targets: The SDG-9 index. Applied Energy, 265, Article 114755. https://doi.org/10.1016/j.apenergy.2020.114755

• Lawrence, M., Homer-Dixon, T., Janzwood, S., Rockström, J., Renn, O., & Donges, J. F. (2024). Global polycrisis: The causal mechanisms of crisis entanglement. Global Sustainability, 7, Article e6. https://doi.org/10.1017/sus.2024.1

• Lusseau, D., & Mancini, F. (2019). Income-based variation in Sustainable Development Goal interaction networks. Nature Sustainability, 2(3), 242-247. https://doi.org/10.1038/s41893-019-0231-4

• MacFeely, S. (2020). Measuring the Sustainable Development Goal indicators: An unprecedented statistical challenge. Journal of Official Statistics, 36(2), 361-378. https://doi.org/10.2478/jos-2020-0019

• Monsod, T. M. C., Majadillas, M. A., & Gochoco-Bautista, M. S. (2023). Unlocking the flow of finance for climate adaptation: Estimates of 'fiscal space' in climate-vulnerable developing countries. Climate Policy, 23(6), 735-746. https://doi.org/10.1080/14693062.2023.2224281

• Moyer, J. D., & Hedden, S. (2020). Are we on the right path to achieve the sustainable development goals? World Development, 127, Article 104749. https://doi.org/10.1016/j.worlddev.2019.104749

• Naidoo, R., & Fisher, B. (2020). Reset Sustainable Development Goals for a pandemic world. Nature, 583(7815), 198-201. https://doi.org/10.1038/d41586-020-01999-x

• Organisation for Economic Co-operation and Development. (2025). Global outlook on financing for sustainable development 2025: Towards a more resilient and inclusive architecture. OECD Publishing. https://doi.org/10.1787/753d5368-en

• Pradhan, P., Weitz, N., Daioglou, V., Abrahão, G. M., Allen, C., Ambrósio, G., Arp, R., Asif, F., Bennich, T., Benton, T. G., Biermann, F., Carlsen, H., Hubacek, K., Leininger, J., Lotze-Campen, H., Pedercini, M., Prell, C., van Vuuren, D., Warchold, A., ... Zimm, C. (2024). Three foci at the science-policy interface for systemic Sustainable Development Goal acceleration. Nature Communications, 15(1), Article 8600. https://doi.org/10.1038/s41467-024-52926-x

• Sachs, J. D., Schmidt-Traub, G., Mazzucato, M., Messner, D., Nakicenovic, N., & Rockström, J. (2019). Six transformations to achieve the Sustainable Development Goals. Nature Sustainability, 2(9), 805-814. https://doi.org/10.1038/s41893-019-0352-9

• Sénit, C.-A. (2020). Leaving no one behind? The influence of civil society participation on the Sustainable Development Goals. Environment and Planning C: Politics and Space, 38(4), 693-712. https://doi.org/10.1177/2399654419884330

• Sénit, C.-A., Biermann, F., & Kalfagianni, A. (2017). The representativeness of global deliberation: A critical assessment of civil society consultations for sustainable development. Global Policy, 8(1), 62-72. https://doi.org/10.1111/1758-5899.12371

• Shulla, K., Voigt, B.-F., Cibian, S., Scandone, G., Martinez, E., Nelkovski, F., & Salehi, P. (2021). Effects of COVID-19 on the Sustainable Development Goals (SDGs). Discover Sustainability, 2(1), Article 15. https://doi.org/10.1007/s43621-021-00026-x

• United Nations. (2025). The Sustainable Development Goals report 2025. United Nations Department of Economic and Social Affairs. https://unstats.un.org/sdgs/report/2025/

• United Nations Conference on Trade and Development. (2023). World Investment Report 2023: Investing in sustainable energy for all. United Nations. https://unctad.org/publication/world-investment-report-2023

• Xing, L., Lu, Y., Wang, J., Chen, W., Liu, L., Pradhan, P., Bryan, B. A., Moallemi, E. A., Gao, L., Schaubroeck, T., Carrasco, L. R., Banik, D., Xu, Z., Li, Y., Zhang, X., Ding, Z., Liu, J., Zeng, S., & Wu, J. (2025). Country-specific progress toward the Sustainable Development Goals: Past, present, and prospects. Proceedings of the National Academy of Sciences, 122(52), Article e2524299122. https://doi.org/10.1073/pnas.2524299122

Hashtags:

#SDGs #2030Agenda #GlobalGoals #SDG1 #SDG8 #SDG9 #SDG10 #SDG17 #SustainableDevelopmentGoals #PolicyCoherence #SDGIndicators #DevelopmentFinance #GlobalGovernance #U7YJournal #U7Y #AcademicResearch