107159 research outputs found
Sort by
Data and Code: Trade Shocks and Credit Reallocation
No full textThis paper identifies a credit-supply contraction that arises endogenously after trade liberalization. Banks with loan portfolios concentrated in sectors exposed to competition from China face an increase in non-performing loans after China's entry into the World Trade Organization. As a result, they reduce the supply of credit to firms, irrespective of the firm's sector of operation. This cut in credit translates into lower employment, investment, and output. Through this mechanism, the financial channel amplifies the shock to firms already hit by import competition from China and passes it on to firms in sectors expected to expand upon trade liberalization
Data and Code for: "The Making of Civic Virtues: A School-Based Experiment in Three Countries"
No full textThis dataset comes from a RCT on civic education. The goal of the study is to examine the effects of the Active citizenship (ACT) program on students’ civic skills and democratic engagement as well as on teachers’ teaching practice. The program, which primarily consists in providing teachers with a 2-days training and with continuous guidance over the school year to help them implement a
citizenship project in their classroom, was implemented in France, Greece and Spain in 2018-2019
Replication Package for "Safe Spaces for Teenage Girls in a Time of Crisis"
No full textThis package contains the data, programs and instructions to replicate the manuscript "Safe Spaces for Teenage Girls in a Time of Crisis," by Oriana Bandiera, Niklas Buehren, Markus Goldstein, Imran Rasul and Andrea Smurra forthcoming at JEEA
An international comparison of longitudinal health data collected on long COVID in nine high income countries: a qualitative data analysis
No full textAbstract Background Long coronavirus disease (COVID) presents a significant health challenge. Long-term monitoring is critical to support understanding of the condition, service planning and evaluation. We sought to identify and examine longitudinal health data collected on long COVID to inform potential decisions in England regarding the rationale for data collection, the data collected, the sources from which data were collected and the methods used for collection. Methods We included datasets in high-income countries that experienced similar coronavirus disease 2019 (COVID-19) waves to England pre-vaccine rollout. Relevant datasets were identified through literature searches, the authors’ networks and participants’ recommendations. We undertook semi-structured interviews with individuals involved in the development and running of the datasets. We held a focus group discussion with representatives of three long COVID patient organisations to capture the perspective of those with long COVID. Emergent findings were tested in a workshop with country interviewees. Results We analysed 17 datasets from nine countries (Belgium, Canada, Germany, Italy, the Netherlands, New Zealand, Sweden, Switzerland and the United Kingdom). Datasets sampled different populations, used different data collection tools and measured different outcomes, reflecting different priorities. Most data collection was research (rather than health care system)-funded and time-limited. For datasets linked to specialist services, there was uncertainty surrounding how long these would continue. Definitions of long COVID varied. Patient representatives’ favoured self-identification, given challenges in accessing care and receiving a diagnosis; New Zealand’s long COVID registry was the only example identified using this approach. Post-exertion malaise, identified by patients as a critical outcome, was absent from all datasets. The lack of patient-reported outcome measures (PROMs) was highlighted as a limitation of datasets reliant on routine health data, although some had developed mechanisms to extend data collection using patient surveys. Conclusions Addressing research questions related to the management of long COVID requires diverse data sources that capture different populations with long COVID over the long-term. No country examined has developed a comprehensive long-term data system for long COVID, and, in many settings, data collection is ending leaving a gap. There is no obvious model for England or other countries to follow, assuming there remains sufficient policy interest in establishing a long-term long COVID patient registry
Supplementary materials and code
No full textPre-publication manuscript and supplementary materials (including code) for "Why do people cooperate with the police and criminal courts? A test of procedural justice theory in 30 countries" (paper being published in Criminology
Code for: "Five Facts about MPCs: Evidence from a Randomized Experiment"
No full textWe present five facts from an experiment on the marginal propensity to consume (MPC) out of transitory transfers: (1) the one-month MPC on a cash-like transfer is 23%; (2) it is substantially higher (61%) on a transfer administered via a card where the remaining funds expire after three weeks, inconsistent with money fungibility; (3) the consumption response is concentrated in the first three weeks; (4) MPCs vary with household characteristics, but are high even for the liquid-wealthy; (5) the unconditional MPC distribution exhibits large variation. Our findings inform the design of stimulus policies and pose challenges to existing macroeconomic models
Limits of ChatGPT's Conversational Pragmatics in a Turing Test About Ethics, Commonsense, and Cultural Sensitivity
No full textDoes ChatGPT deliver its explicit claim to be culturally sensitive and its implicit claim to be a friendly digital person when conversing with human users? These claims are investigated from the perspective of linguistic pragmatics, particularly Grice's cooperative principle in communication. Following the pattern of real-life communication, turn-taking conversations reveal limitations in the LLM's grasp of the entire contextual setting described in the prompt. The prompts included ethical issues, a hiking adventure, geographical orientation and bodily movement. For cultural sensitivity the prompts came from a Pakistani Muslim in English language, from a Hindu in English, and from a Chinese in Chinese language. The issues were deeply cultural issues involving feelings and affects. Qualitative analysis of the conversation pragmatics showed that ChatGPT is often unable to conduct conversations according to the pragmatic principles of quantity, reliable quality, remaining in focus, and being clear in expression. We conclude that ChatGPT should not be presented as a global LLM but be subdivided into several culture-specific modules.
Data are stored and provided as a docx file.
The file contains 8 exchanges and each one comprises the following:
Metadata
Date: the date of the exchange
Domain: a thematic category among ethics, common sense, and cultural sensitivity
Participant: AuthorsTeam
Interviewer: ChatGPT-4o
Question-answer pairs between the participant and the interviewe
Corpus of Serbian Online News Articles and Reader Comments on the Russia–Ukraine War (February–August 2022)
No full textThis dataset contains a corpus of 4,201 news articles and associated reader comments collected from four major Serbian online newspapers — Novosti, Politika, Blic, and Danas — during the first six months of Russia’s full-scale invasion of Ukraine (24 February – 31 August 2022). The corpus includes articles explicitly mentioning Russia or Ukraine and at least one associated reader comment. It covers outlets across the ideological spectrum, from nationalist right to liberal left, and was used to examine how debates over Serbia’s own war legacies shaped public responses to reports of war atrocities in Ukraine. The dataset has been cleaned and prepared for both quantitative and qualitative text analysis. It serves as a resource for studying online deliberation, transitional justice, vicarious trauma, and denial in post-conflict societies.(2025-08-24)
Funding
Lanabi La Lova and Denisa Kostovicova gratefully acknowledge funding for this research from the European Research Council (ERC) Consolidator Grant “Justice Interactions and Peacebuilding: From Static to Dynamic Discourses across National, Ethnic, Gender and Age Groups” (JUSTINT), No. 772354
Replication Data for: Bribery, Inequality and Reciprocity - Experimental Evidence from a Novel Allocation Game
No full textThis is the dataset used to conduct all analyses present in the paper "Bribery, Inequality and Reciprocity - Experimental Evidence from a Novel Allocation Game
