Recent crises – from the Global Financial Crisis to COVID-19 and the 2022-2023 energy crisis – have highlighted many governments’ limited informational capacity, that is, their inability to swiftly gather and analyse data for effective crisis responses. These crises have also challenged assumptions that weak state capacity mainly affects low- to middle-income countries. Indeed, many rich liberal democracies increasingly resemble analogue titans in a digital age. In this column we therefore explore why informational challenges lie at the heart of crisis management, affecting equity, efficiency, and the overall effectiveness of government interventions. Specifically, we use theory and simulations to compare the measures adopted by Germany and the UK in response to the energy crisis caused by Russia’s invasion of Ukraine.
Beyond fiscal and legal capacity: Incorporating the informational dimension
State capacity – the ability of governments to implement policies – is typically conceptualised as consisting of two components: fiscal and legal capacity (Besley and Persson 2009, 2011, Dann et al. 2021, Suryanarayan 2024). Fiscal capacity refers to governments' ability to levy taxes (Bergeron et al. 2024, Martin 2023, Queralt 2015), while legal capacity pertains to judicial independence and the rule of law (Fukuyama 2011). Drawing on the work of James C. Scott (1999) and other historical political economy research (Ansell and Lindvall 2020; Brambor et al. 2020), scholars have identified a third component – informational capacity.
Yet historical perspectives shed no light on contemporary governments' limitations in gathering, processing, and analysing data. These limitations include outdated IT systems, fragmented interdepartmental data sharing, limited analytical expertise, and restrictive data-use laws – which we term the state's ‘informational boundaries’. In a recent paper (Fetzer et al. 2024), we develop an empirically operationalisable framework to examine the effects of relaxing policymakers' informational constraints.
The UK and German policy responses
To understand our theoretical set-up, it is instructive to juxtapose the UK and German responses to the 2022-2023 energy crisis (Arregui 2022). The UK's response involved a lump-sum payment of £400 per household in March 2022, followed by a broad price cap on energy bills (the Energy Price Guarantee, or EPG). While this provided immediate relief, the policy muted price signals, reducing incentives for households to conserve energy and thus reduce emissions (fossil fuel subsidies). The untargeted nature of the subsidy had regressive distributional effects, disproportionately benefitting higher-income households (Alpino et al. 2025, Levell et al. 2024). Unlike Germany, the UK’s choice reflected not a lack of high-quality data, but rather – as we show below – limited governmental capacity to use available data, stemming from austerity-driven personnel cuts and a longstanding neglect of statistical literacy in civil service recruitment (Feld and Fetzer 2024, Freedman 2024).
Germany pursued a more targeted approach, offering lump-sum transfers tied to historical household energy consumption to preserve price signals and incentivise conservation (ExpertInnen-Kommission Gas und Wärme 2022). Implementation, however, proved challenging, hindered by insufficiently granular data and bureaucratic complexity stemming from decentralised governance. Both cases highlight how governmental informational capacity fundamentally shapes crisis responses.
Modelling framework
Our analysis centres on a theoretical framework illustrating how informational capacity shapes policymakers’ optimal choices. In the model, a policymaker allocates a fixed budget between a lump-sum transfer and a price subsidy. The policymaker observes consumer types – rich and poor – imperfectly, potentially resulting in misdirected transfers (e.g. rich households receiving support), reflecting constraints on informational capacity. Since, as we document, high-income households consume more energy, each pound spent on the subsidy disproportionately benefits them. The policymaker also has distributional preferences, placing greater weight on the welfare of either poor or rich consumers.
The framework yields three key results. First, with full information, policymakers can achieve distributional goals solely through lump-sum transfers, avoiding distortionary subsidies – a standard result in the theory of commodity taxation. Second, as informational uncertainty rises, governments shift from targeted transfers toward subsidies, as it becomes harder to identify who genuinely needs support. Because past consumption is easier to observe, setting subsidy quotas is more feasible than targeting transfers. Thus, policymakers can better achieve distributional objectives through subsidies, despite their inefficiency in distorting price signals.
Third, under maximal uncertainty, policymakers with regressive preferences (favouring the rich) fare better than progressive ones. Since subsidies benefit high-income households more due to higher consumption, ‘regressive policymakers’ find inefficiency less costly. In contrast, ‘progressive policymakers’ incur greater welfare losses, as the inefficient allocation undermines their redistributive aims.
Empirical evidence and simulations
Using detailed UK data covering around 60% of properties – including energy consumption, property characteristics, and other socioeconomic variables – we simulate nearly 57,000 alternative responses to the energy crisis. These include an untargeted price subsidy (like the UK’s EPG) and a menu of two-tier tariffs combined with lump-sum transfers. The two-tier structure subsidises a baseline quota, with excess consumption subject to market prices. Informational capacity is proxied by the dimensionality of the vector of observable household characteristics used for targeting transfers. Higher dimensionality implies greater informational capacity and reduced privacy.
We measure efficiency by the proportion of households facing market prices, and equity by the share benefiting relative to the implemented policy. Distributional objectives assume the incumbent Conservative Party selects policies generating a positive correlation between household transfers and Conservative support. Given the positive correlation between energy use and Conservative support, this implies regressive preferences – favouring richer households. All simulated policies are ex-ante fiscally neutral (equal in cost to the EPG) and implementable using publicly available data.
Simulation results
The simulations show that when policymakers can target transfers using a richer set of observable characteristics – that is, when they have greater informational capacity – they can design policies that are both more equitable and more efficient, while still achieving their distributional objectives.
Figure 1 illustrates these observations. Panel A shows that most alternatives preserve a strong correlation between subsidised bills and Conservative support, similar to the actual EPG. This suggests that regressive distributional goals could still have been met under more efficient and equitable policies. Panel B highlights that alternative policies typically maintain market signals for a much larger share of consumption—more in line with Germany’s targeted approach. Panel C addresses the privacy dimension. Even when targeting uses high-dimensional observables, households are generally grouped into sufficiently coarse blocks, avoiding quasi-individualised transfers and preserving privacy. Panel D reveals substantial welfare improvements: a majority of households could have been better off under alternative policies, underscoring the gains achievable through enhanced informational capacity.
Figure 1 Characterisation of the empirical distribution of fiscally neutral two-tier tariff alternatives vis-à-vis equivalents of the UK and German policy responses, respectively
Notes: These figures show the empirical distribution from a broad evaluation of fiscally neutral two-tier block tariffs – assessed across several metrics – relative to the UK’s EPG. Panel A shows the correlation between net consumer bills (post-subsidy) and Conservative vote share. Panel B depicts the share of consumption facing market prices. Panel C shows the distribution of tariff blocks with fewer than 10 households. Panel D presents the share of households that, all else equal, would be better off relative to the EPG or Germany’s policy.
In addition to the reasons discussed above for the inability to use publicly available data, Figure 2 suggests that correcting this lack of informational capacity may run into political challenges. The figure shows that the main beneficiaries from the capacity deficit evident in the EPG were Conservative supporters in the top 10% of the income distribution (Panel B), rather than the average Tory voter (Panel A). To the extent that wealthy Tory supporters exercise outsized influence on political decisions, increasing informational capacity will be a politically dicey proposition.
Figure 2 Empirical distribution of the average degree of political targeting and partisanship for the top 10 income percentile
Notes: Panel A presents the empirical distribution of the different correlation coefficients, capturing the correlation between the net-of-transfer energy bills faced by consumers and whether an individual is supporting the Conservative Party. Panel B presents the same relationship as Panel A, but focuses on the correlation among households in the top 10% of the income distribution.
Overall, the simulations highlight the substantial benefits of investing in data infrastructure and analytical capabilities. Improved informational capacity enables the design of crisis-response policies that are not only fiscally neutral, but also more cost-effective, equitable, and similarly politically feasible.
Conclusion and lessons for the future: Rethinking government use of information
The experiences of the UK and Germany during the energy crisis offer broader lessons about the importance of informational capacity for liberal democracies. Unlike digitally savvy autocracies that readily infringe on citizens’ rights (Gohdes 2024, Guriev and Treisman 2022), liberal democracies must protect their citizens’ privacy while realising the gains from ramping up their informational capacity. Forgoing these gains risks costly and ineffective policies that may end up undermining the ‘output legitimacy’ (Scharpf 1997) of liberal democracies (i.e. the willingness of citizens to comply with laws because the system, on average, delivers desirable outcomes (Schularick 2021)). This is especially troubling amid rising right-wing populism, which thrives on low trust in government and narratives of state failure.
Balancing privacy, informational capacity, and output legitimacy is difficult but feasible. Achieving this balance, however, requires more policy experimentation and public discourse than is currently the case. These discussions should address both the technological tools – such as homomorphic encryption, decentralised data storage, and differential privacy – and the institutional arrangements, including hiring practices that reward statistical literacy, needed to give liberal democracy a much-needed digital overhaul. Our findings suggest that the cost of avoiding this discourse is high – both economically and politically.
References
Alpino, M, E Ciapanna, L Citino, and G Rovigatt (2025), “Navigating the Electric Storm: Assessing Policy Responses to Europe’s Energy Shock”, CEPR Discussion Paper 19981.
Ansell, B W and J Lindvall (2020), Inward Conquest: The Political Origins of Modern Public Services, Cambridge University Press.
Arregui, N (2022), “Targeted, Implementable, and Practical Energy Relief Measures for Households in Europe”, IMF Working Papers 2022/262).
Bergeron, A, G Tourek, and J L Weigel (2024), “The State Capacity Ceiling on Tax Rates: Evidence From Randomized Tax Abatements in the DRC”, Econometrica 92(4): 1163–93.
Besley, T and T Persson (2009), “The Origins of State Capacity: Property Rights, Taxation, and Politics”, American Economic Review 99(4): 1218–44.
Besley, T and T Persson (2011), Pillars of Prosperity: The Political Economics of Development Clusters, Princeton University Press.
Brambor, T, A Goenaga, J Lindvall, and J Teorell (2020), “The Lay of the Land: Information Capacity and the Modern State”, Comparative Political Studies 53(2): 175–213.
Dann, C, T Persson, and T Besley (2021), “State Capacity and Development Clusters”, VoxEU.org, 18 June.
ExpertInnen-Kommission Gas und Wärme (2022), Sicher Durch Den Winter: Abschlussbericht, BMWK.
Feld, I and T Fetzer (2024), "Performative State Capacity and Climate (In)Action”, ECONtribute Discussion Paper.
Fetzer, T, C Shaw, and J Edenhofer (2024), “Informational Boundaries of the State”, CESifo Working Paper.
Freedman, S (2024), Failed State: Why Nothing Works and How We Fix It, Macmillan.
Fukuyama, F (2011), The Origins of Political Order: From Prehuman Times to the French Revolution, Profile Books.
Gohdes, A R (2024), Repression in the Digital Age: Surveillance, Censorship, and the Dynamics of State Violence, Oxford University Press.
Guriev, S, and D Treisman (2022), Spin Dictators: The Changing Face of Tyranny in the 21st Century, Princeton University Press.
Levell, P, M O’Connell, and K Smith (2024), “Distributional Effects of the European Energy Crisis”, working paper.
Martin, L E S (2023), Strategic Taxation: Fiscal Capacity and Accountability in African States, Oxford University Press.
Queralt, D (2015), “From Mercantilism to Free Trade: A History of Fiscal Capacity Building”, Quarterly Journal of Political Science 10(2): 221–73.
Scharpf, F W (1997), “Economic Integration, Democracy and the Welfare State”, Journal of European Public Policy 4(1): 18–36.
Schularick, M (2021), Der entzauberte Staat: Was Deutschland aus der Pandemie lernen muss, C.H.Beck.
Scott, J C (1999), Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed, Yale University Press.
Suryanarayan, P (2024), “Endogenous State Capacity”, Annual Review of Political Science 27: 223–43.
