How effectively the machinery of governance enforces laws, regulates economic activity and provides public services is a key issue around the world. The Covid-19 pandemic and other policy challenges highlight wide variations in state capacity among countries – and the growing strain on public trust.
This article is part of an upcoming collection on state capacity and industrial strategy by the Policy Hub for the Huth Initiative for a New Political Economy.
Having a capable state matters for lots of different outcomes. But the capacity of the state to realise its objectives is often taken for granted.
Its absence is felt most strongly in fragile, conflict-affected and low-income countries stuck in a poverty trap. But the ability of the state to enforce laws, regulate economic activity and provide public services is also significant in higher-income countries, especially in view of the unique policy challenges of the 21st century.
Unpacking state capacity, measuring its various dimensions and determining its drivers remain key issues for public policy as well as development. Even so, it was not a central concern of policy paradigms in the recent past.
For example, the Washington Consensus had little to say about the internal functioning of the state beyond recognising the importance of secure property rights and getting prices and policies right. Subsequent experience and work (such as the London Consensus, 2025) have shown the limitations of this approach, shifting the focus towards the ability of the state to adopt, implement and course-correct.
Research shows that an effective state is an essential complement to economic growth and development. Indeed, a study from 2021 uses bureaucratic effectiveness as a measure of state capacity. The authors find that bureaucratic quality and economic development (measured by GDP per capita) are strongly correlated, and that improvements in bureaucratic capacity are associated with growth.
The past decades have seen significant advances towards understanding the determinants of state capacity, but questions remain. Covid-19 and other big new policy challenges highlight the importance of rethinking our approach to capacity, not least as the response and performance of different countries in the pandemic did not, in many cases, match the various measures and indicators of their preparedness.
Many policy challenges today do not fit standard institutional responses. Examples include enhancing learning and health outcomes rather than improving access to education and healthcare, better preparedness for new pandemics, addressing issues around climate change and nature, and increasing economic security and strategic resilience.
Instead, they require complementarity between new forms of enforcement capacity, with novel forms of citizen compliance. The necessary compliance with policies, taxes and laws is affected by the level of public trust, which is itself driven by beliefs about whether the government is acting in the public interest.
Public trust in many parts of the world, including established democracies, is under growing strain. For example, only four in ten people in OECD countries trust their national government or believe that their government uses the best available evidence when making a decision (OECD, 2024). This is a damning statistic.
Lessons from Covid-19: how does state capacity vary across countries?
The capacity of the state to perform its functions varies hugely across countries and functions. It can also wax and wane over time. In this sense, there is no single, unified capacity.
Covid-19 brought this variability into sharp focus. Different states had different experiences as they tried to respond to the crisis, and citizen compliance, lockdown effectiveness, excess mortality and vaccine rollouts all varied from country to country.
In recent work, we illustrate this variation (what economists call heterogeneity) by looking at data on income per capita, educational outcomes (measured using scores for learning based on the OECD’s Programme for International Student Assessment, PISA) and Covid-19 mortality.
Figure 1 shows the relationship between income and educational outcomes, using pupils’ mathematics scores. The chart shows a positive relationship, with some notable over- and underperformers. For example, Estonian pupils outscore their American peers by the equivalent of over two years of schooling, despite the Baltic state’s GDP per capita being 40% lower, Note that on average, what pupils learn over a school year corresponds to about 20 score points in PISA (though this estimate is based on 15-year-olds).
Figure 1. PISA mathematics scores by GDP per capita, 2022
Source: OECD PISA 2022, World Bank.
Note: Ireland, Norway, Singapore and Qatar excluded.
Figure 2. Covid-19 excess mortality by GDP per capita, 2020-22
Source: Our World in Data Covid-19 dataset.
Note: Excess mortality is measured using the P-score, which expresses the percentage difference between observed deaths and the expected number of deaths based on pre-pandemic trends (2015–19 baseline). Higher values indicate more deaths than expected.
The relationship between income and Covid-19 outcomes is less clear – though richer countries generally fared better (see Figure 2). One striking comparison is Mongolia and Peru. The two countries have very similar GDP per capita levels (approximately $15-16,000), yet experienced vastly different Covid-19 outcomes. Peru saw 45% more deaths than expected based on pre-pandemic trends, while deaths in Mongolia were ‘only’ 1.4% above expected.
The same pattern holds at higher income levels. Kuwait (GDP per capita of $46,000) recorded 32% more deaths than expected, while New Zealand ($49,000) saw slightly fewer deaths than expected (-0.3%).
Figure 3. PISA mathematics scores in top- and bottom-scoring countries with differences in equivalent years of schooling, 2022
Source: Author’s calculations, OECD PISA 2022.
Note: The 15 top and bottom-scoring countries are ranked by average (mean) score. A difference of 20 PISA points is treated as approximately equivalent to one year of schooling, based on OECD assessments of 15-year-olds. Comparisons involving longer time horizons or larger gaps between countries should be interpreted with caution.
Looking at the distribution of educational outcomes, the gap between the highest and lowest performing countries is 239 points (see Figure 3). Singapore (575) is at the top and Cambodia takes the bottom spot (336). Using the OECD assessments of 15-year-olds implies that this point gap is comparable to 12 years of schooling (although these equivalences are more illustrative benchmarks of learning gaps rather than direct measures of time spent in education).
Gaps also exist within national cohorts. The average gap between the 90th and 10th percentile pupil (the top 10% of performers versus the bottom 10%) within the same country is 219 points (or 11 years) across all countries, rising to 12 years for OECD countries. In other words, the average educational inequality within countries is almost as large as the inequality between them.
Again, Estonia offers a notable example. It achieves one of the highest average (mean) scores (510) while maintaining a relatively narrow gap between its best and worst performing pupils: 219 points (or 11 years of schooling).
The Netherlands tells the opposite story. Despite a GDP per capita 60% higher than Estonia’s ($79,000 compared with $49,000), it scores lower on average (493) and has the widest within-country gap (282 points). Within the same education system, the gap between the highest and lowest performing pupils is comparable to 14 years of schooling. These figures illustrate the point that income alone does not determine the quality or equity of public service delivery.
Figure 4. Government effectiveness scores for select countries, 2005-24
Source: World Bank Governance Indicators 2025.
Note: Government Effectiveness (GE) score ranges from approximately -2.5 to 2.5, where higher values indicate more effective government.
Capturing state capacity is hard, but measures like the World Bank’s Government Effectiveness (GE) can serve as sensible proxies (see Figure 4). The GE score captures perceptions of the quality of public services, the civil service, policy formulation and implementation, and the credibility of a government’s decisions.
The measure shows that across income levels, different countries have seen capacity rise and fall over last 20 years. For example, since 2005, the UK has seen a decline in its GE score of 0.65, which is greater than the 0.54 decline in Afghanistan over the same period.
Although GE alone cannot capture various dimensions of state capacity, it seems to capture something meaningful. It shows a stronger association with Covid-19 excess mortality than GDP, with an R² value (a measure of correlation between zero and one) of 0.29 compared with an R² of 0.15 for GDP per capita.
Figure 5 shows that countries with more effective governments, as measured by GE, tended to have lower excess mortality during Covid-19. A couple of notable outliers here are Mexico and Ukraine. Both score relatively poorly on GE, yet Mexico recorded excess mortality 12% above expected levels during Covid-19, while Ukraine experienced mortality rates 5% below expected levels.
At the other end of the distribution, New Zealand’s excess mortality remained 8% below expected despite already ranking among the highest performers on GE.
Figure 5. Covid-19 excess mortality by Government Effectiveness score, 2020-22
Source: Author’s calculations, Our World in Data Covid-19 dataset; World Bank Governance Indicators.
Figure 6 shows the relationship between income, health and educational outcomes. Not unexpectedly, higher-income economies tend to cluster in the higher PISA scores and lower Covid-19 excess mortality quadrant. While richer countries generally achieve better educational and health outcomes, this relationship is not uniform and is not guaranteed.
Figure 6. PISA mathematics score and Covid-19 pandemic excess mortality by GDP per capita
Source: OECD PISA 2022, Our World in Data Covid-19 dataset
Note: Bubble size denotes population.
What are the ways to build an effective state?
There are many dimensions in which state capacity and government effectiveness vary – across and within countries, by income and by functions. Much remains to be done to demonstrate this heterogeneity better, to explain its determinants and to identify how it can be strengthened.
Our understanding of the black box of state capacity and effectiveness has been much enriched by research on development and political economy. This work has provided insights on how incentives and ‘selection mechanisms’ operate. Other research has highlighted the importance of motivated public officials and of harnessing their ‘pro-social’ motivations to build mission-driven organisations.
Recent work in political economy (in particular, studies by Besley and Persson and Acemoglu and Robinson) has deepened our understanding of how political forces affect incentives to invest in state capacity, and how this is fundamentally a political challenge and not just a technical one.
But while barriers to state effectiveness like information, monitoring or political economy issues can often be clearly identified, the pathway through which improvement in state capacity might occur is often unclear.
In such cases, the focus should not be on implementing a particular reform but on building processes and environments in which constraints on better implementation can be identified and addressed. Such adaptive approaches to building state capacity distinguish between formal (mimicked) and real capability, and view it as an emergent contingent capacity as opposed to a ‘copy-ready’ capacity.
What are the gaps in state capacity research?
Despite progress, developing appropriate measures of state capacity is challenging. Measuring state capacity in its various dimensions and identifying the drivers of capacity in weakly institutionalised environments remain foundational questions.
Availability of new data and new measurement technologies (for example, remote sensing data or the use of artificial intelligence), along with creative uses of administrative data and subjective measures, can potentially help to move beyond coarse measures and unblock some hurdles in this research space.
The goal is to uncover the causal drivers of building capable states. When conducted in collaboration with governments, and when frictions in the pathways to adoption and implementation are also addressed, such work can also help to build greater capacity on the ground.
In future, such work needs to go beyond economics to make better connections across disciplines in a way that captures the richness of motivation, the role of social norms and identity.
This is even more relevant given the emergence of new policy challenges. Whether state systems and bureaucracies can innovate and adapt to future challenges like new pandemics, climate change, political and social polarisation, and economic insecurity remains an important question.
Perhaps the biggest gap in our understanding of the drivers of state capacity is in how best to design and roll out system-wide reforms and find ways to evaluate policies ‘at scale’. This is particularly hard, which means that credible evidence on system-level questions remains scarce.
What next for state capacity research?
Researchers should shift the focus on their work on state effectiveness away from ‘personnel’ economics and towards more organisational and political economy-based discussions. Going beyond traditional economics is critical to unlocking ideas based on frontier research that can shape the world of policy and practice.
This is particularly salient in the case of fragile states in the effort to build state legitimacy and effectiveness, and the role of international bodies in building and not undermining domestic capacity. This is also true for the future of liberal democracies and the threat from the rise of populism.
Building state capacity is possible and constitutes the major challenge of our age. It is an essential complement to all other objectives that we all want to achieve. This requires building on the current body of political economy research, integrating results from rigorous microeconomic studies as well as insights from adaptive approaches to broader macroeconomic models. This will help to bring about state reforms that can have a significant positive impact on human wellbeing.
Where can I find out more?
- The London Consensus: economic principles for the 21st century. See the chapters on political institutions and state capacity for accessible summaries of the emerging principles and available evidence.
- VoxDev Literature Review: an accessible synthesis of work on bureaucracy.
- VoxDev: summaries of the latest research in development economics.
- Harvard Building State Capability.
Who are experts on this question?
- Timothy Besley (LSE)
- Torsten Persson (IIES)
- Daron Acemoglu (MIT)
- Jim Robinson (Chicago)
- Lant Pritchett (LSE)
- Guo Xu (UC Berkeley)
- Matt Andrews (HKS)
- Dan Honig (Georgetown)
For political economy frameworks, see work by Timothy Besley (LSE), Torsten Persson (IIES), Daron Acemoglu (MIT) and Jim Robinson (Chicago), and for new empirical work, follow Guo Xu. For unpacking the details of building state capacity and organisational effectiveness, see work by Lant Pritchett (LSE), Matt Andrews (HKS), Dan Honig (Georgetown) and others.
