Next Release

Does deskilling cause unrest? I study this question in the context of the First Industrial Revolution in England, using parish-level data on the adoption of skill-saving machines drawn from a corpus of newspaper advertisements. Reduced-form analysis leveraging an event study research design finds a positive causal effect of skill-saving mechanisation on unrest intensity between 1760 and 1814. Heterogeneity analysis of peaceful collective bargaining and petitioning by employers and workers identifies the weakening of labour market institutions as a plausible mechanism, attributable to the imbalance of lobbying power between winners and losers of disruptive technological change.

Does elite conflict cause franchise extension? I study this question in the context of the 1832 Great Reform Act in England, using surviving poll books and parish records on the formation of disenfranchising select vestries. Reduced-form analysis leveraging a difference-in-differences research design finds that political elites residing in parishes with democratic backsliding increased their electoral support for the reformist Whig party between the 1830 and 1831 general elections. I attribute the research findings to the outbreak of the Swing Riots, which occurred with greater intensity in select vestry parishes, providing an elite-conflict interpretation of the theory of revolutionary threat.

Does welfare retrenchment cause social unrest? I study this question using parish-level expenditure on poor relief and data on social unrest in post-Napoleonic England, drawing on the 1818-9 welfare reforms to motivate instrumental variables, regression discontinuity, and event study research designs. The reduced-form analysis finds a causal effect of welfare retrenchment on social unrest between 1816 and 1834, attributable to an exacerbation of relative deprivation that stems from comparisons with higher-expending parishes. Simulations from a counterfactual analysis assuming greater accountability for these grievances highlight the importance of equitable social welfare provision in minimising the political costs of austerity.

This dataset covers the universe of speeches made in the Egyptian House of Representatives (Majlis al-Nuwwab) in 1924-1952. The dataset was constructed by the data creators from high-quality images of the Arabic-language minutes of the daily meetings of parliament created and made available online by the Institute of Oriental Culture, University of Tokyo, Japan. The data construction consisted of three stages: (1) image segmentation into text block images, (2) implementing Optical Character Recognition (OCR) on the corpus of text block images using Google Document AI, and (3) extraction of speeches and speaker names using OpenAI. The dataset is at the speech level. It provides the following information for each speech: text body of the speech, speaker's full name, politician ID (including whether speaker is member of parliament or minister), date of meeting, parliamentary session, and parliamentary cycle. While most of these speeches are made by members of parliament and ministers, other speakers include high-level bureaucrats and members of the public who presented petitions to the parliament.