107159 research outputs found
Sort by
Addressing the learning crisis: an emergent consensus
No full textSuccess can make a previous consensus not so much wrong as just irrelevant. The Washington Consensus joined in a broader consensus that governments need to spend on education in order to reach universal schooling to create human capital. But ‘spend to expand access’ has been so successful there is less and less space for additional improvements in education outcomes – the skills and competencies children need to acquire in school – through ‘access’. Global, national, and local actors agree on the need to increasingly focus on improving learning outcomes. Moreover, there is an emergent consensus that improving learning will require much more than just ‘more spend’ and that a substantial re-alignment of education systems from ‘expansion of access’ to ‘increased learning’ is needed. And, while there is not yet a consensus on the granular details (and may never be as success tends to be home-grown and adapted to context), there is increasing agreement around a set of principles that will drive sustained gains in improving learning outcomes
State capacity
No full textThe Washington Consensus had little to say regarding the internal functioning of the state beyond recognising the importance of securing property rights and general policies, such as ‘broadening the tax base’. The past decades have seen significant advances towards understanding the determinants of state capacity. In this chapter we do not attempt to provide universal prescriptions, but identify a few general principles that we believe apply in all countries, irrespective of wealth or level of development. Effective governance requires more than merely setting correct policies; it demands a state’s ability to implement, adapt, and learn over time. Drawing from historical and political economy perspectives, as well as contemporary managerial approaches, we examine the evolution and variation of state capacity, focusing on sectors like tax administration and health. State capacity is a process, not an event; hence medium-term commitment is critical, as is acting in the understanding of long-term constraints and environment. That process should begin with diagnostics of a particular place and goal, not universal prescriptions. Building state capacity involves a continuous process of, adaptation, and systemic learning; empowered agents, citizen-state trust, and iterative improvements are often key components of success. Ultimately, enhancing state capacity requires not only technical solutions but also alignment with political incentives and stakeholder motivations, recognising that a capable state is foundational to enabling citizens and markets to thrive
When self-censorship stifles classroom debate
No full textA spiral of silence can inhibit diverse viewpoints even in diverse classrooms, argues Chen-Ta Sun
From ideology to economy: how Confucianism and the Protestant ethic molded cultural norms, institutions, and divergent paths in Imperial China and early modern Europe
No full textThis essay compares the influence of Confucianism in China and the Protestant ethic in Europe on both formal and informal institutions, examining their role in shaping divergent economic trajectories. Drawing on historical and institutional analysis, this essay integrates insights from economic history, sociology, and political theory. The findings contribute to debates on the cultural origins of the Great Divergence and offer broader insights into how culture interacts with governance structures and economic incentives over the long run. Understanding these historical dynamics is valuable not only for explaining the Great Divergence but also for interpreting contemporary patterns of development, governance, and social trust. In an era where policymakers and international organizations grapple with institutional reform, corruption, and cultural barriers to economic growth, the study highlights the importance of aligning institutional design with prevailing social norms to foster sustainable, inclusive development
Quantifying connectivity: the causal effect of railway accessibility on local industrial economic outcomes, France 1846-1865
No full textFrance’s railway expansion following the Law of 11 June 1842 significantly reshaped nationwide connectivity and economic opportunities. This dissertation investigates the causal impacts of railway access between 1846 and 1861 on city-level industrial development. Using a dataset combining industrial surveys with digitized railway records, it employs a robust Difference-in-Differences approach, leveraging the quasi-exogenous roll-out of the centrally planned ‘étoile de Legrand’ railway network. Empirical results show railway access increased industrial activity primarily extensively: railway-connected cities saw approximately a 20% rise in the number of factories and workers, especially in labour-intensive sectors like textile in Lille and ceramics in Limoges. Yet, intensive effects such as factory size, productivity, and wages remained statistically and economically negligible. Contrary to theoretical predictions from trade and New Economic Geography models, capital-intensive sectors, such as metallurgy in Lorraine, did not exhibit statistically significant responsiveness. These findings reframe the role of transport infrastructure from being a deterministic catalyst to being better understood as a conditional enabler. While railways expanded market potential, their short to medium term transformative impact critically depended on complementary institutional frameworks notably financial markets and property rights, technological readiness, and regional contexts. Acknowledging the historical data limitations, this study underscores that transport infrastructure alone is insufficient for structural economic upgrading without the appropriate institutional, technological, and human capital conditions in place at the right time
Doubly robust uncertainty quantification for quantile treatment effects in sequential decision making
No full textWe consider multi-stage sequential decision making, where the treatment at any stage may depend on the subject’s entire treatment and covariate history. We introduce a general framework for doubly robust uncertainty quantification for the quantiles of cumulative outcomes under a sequential treatment rule. While previous studies focused on mean effects, quantile effects offer unique insights into the distributional properties and are more robust for heavytailed outcomes. It is known that, doubly robust inference is significantly more challenging and largely unexplored for quantile treatment effects. More importantly, for mean effects, doubly robust estimation does not ensure doubly robust inference. Our approach first provides a doubly robust estimator for any quantile of interest based on pre-collected data, achieving semi-parametric efficiency. We then propose a novel doubly robust estimator for the asymptotic variance, enabling the construction of a doubly robust confidence interval. To overcome the challenges in parameter-dependent nuisance functions, we leverage deep conditional generative learning techniques. We demonstrate advantages of our approach via both simulation and real data from a short video platform. Additionally, we observe that our proposed approach leads to another mean effect estimator that outperforms existing estimators with heavy-tailed outcomes
Grounding fire: from climate affect to imperfect alliance in La Chiquitania
No full textThe lowland Bolivian region of La Chiquitania, straddling the Amazon and La Plata River basins, has been severely impacted by recent wildfires. Wildfires increasingly act as an extractivist method for deforesting landscapes, facilitating the conversion of land into property and of subsoil, timber, meat, soy, and minerals into commodities slated for international markets. This article explores how wildfires render landscapes disposable not only for capital but also for a set of environmental mediations that can disempower those most afflicted by climate change. For those targeted by these interventions, wildfires and conservation responses reproduce aspects of earlier histories of territorial displacement by colonial Jesuits, Spanish plantation owners, white and mestizo agro-capitalists, and anti-Indigenous conservationists. Building from collaborative research in this region, I ask how these fires elicit new reflections upon a racialized history of land dispossession to which Indigenous people have been subjected. Against presumptions of a shared atmosphere of harm—the myth of universalism underlying what I call climate affect—the allied responses to wildfire in Chiquitos that I discuss foreground the unequal harms of an ongoing but not new climate apocalypse
Beyond average value function in precision medicine: maximum probability-driven reinforcement learning for survival analysis
No full textConstructing multistage optimal decisions for alternating recurrent event data is critically important in medical and healthcare research. Current reinforcement learning (RL) algorithms have only been applied to time-to-event data, with the objective of maximizing expected survival time. However, alternating recurrent event data has a different structure, which motivates us to model the probability and frequency of event occurrences rather than a single terminal outcome. In this paper, we introduce an RL framework specifically designed for alternating recurrent event data. Our goal is to maximize the probability that the duration between consecutive events exceeds a clinically meaningful threshold. To achieve this, we identify a lower bound of this probability, which transforms the problem into maximizing a cumulative sum of log probabilities, thus enabling direct application of standard RL algorithms. We establish the theoretical properties of the resulting optimal policy and demonstrate through numerical experiments that our proposed algorithm yields a larger probability of that the time between events exceeds a critical threshold compared with existing state-of-the-art algorithms
Artificial intelligence in education: computer-assisted learning and AI-guided tutors
No full textArtificial Intelligence (AI) and Computer-Assisted Learning (CAL) offer powerful tools to improve foundational skills and close educational gaps, with evidence showing meaningful gains in student performance, especially in mathematics. Recent advancements in these technologies have generated optimism about their transformative potential in classrooms worldwide. These technologies are increasingly being piloted at scale, reshaping the way teachers deliver content and students engage with material. However, their impact depends less on access to devices and more on how they are integrated into teaching—through curriculum alignment, teacher training, and interactive design that promotes active learning. Without careful implementation, these tools risk widening existing inequalities. Using new evidence from Italy, we show that digital divides in AI adoption persist across schools and regions, reflecting broader social and economic disparities. Our findings suggest that realising the potential of AI in education requires inclusive policies and targeted investment to ensure no student is left behind, and that the benefits of digital innovation are shared equitably
So good, but so far away? The effect of institutional distance on the parent CSR and subsidiary reputation link
No full textMultinational enterprises (MNEs) leverage strategies of Corporate Social Responsibility (CSR) at the parent and subsidiary levels to build a reputation overseas. Nevertheless, institutional distance can weaken this connection in developing host countries, where MNEs face significant institutional voids. We explore the mechanisms through which CSR enhances subsidiary reputation, focusing on how stakeholders in host developing countries perceive CSR signals sent from headquarters. We further explore the moderating role of formal and informal institutional distance in this relationship. Using a panel of MNEs headquartered in developed countries and operating across Latin America, we employ a multi‐stakeholder indicator of the subsidiary reputation based on assessments from key host country stakeholders. The analysis controls for country, corporate, and subsidiary‐level factors, including a variable derived from big data analytics. By examining the cross‐country parent CSR signals and their subsidiary reputation effects, this study advances the international business literature, providing new insights into how institutional distance shapes the local reputational outcomes of parent CSR strategies