Signals on Budget Day: Media Attention and Public Beliefs in Ireland

Ireland is an outlier among European countries in that it has no automatic indexation of tax and welfare parameters. As a result, the annual Budget announcement plays a significant role, both in shaping the government’s discretionary policy priorities and in influencing people’s expectations about the future. In this project, we study Budget Day as a signalling device along two margins. First, we ask how discretionary policy choices across spending areas respond to political economy forces: media attention, organised interest-group pressure, and electoral cycles. Second, we examine whether Budget announcements shift individuals beliefs about redistribution and uncertainty about the future, and whether these responses are heterogeneous by predicted exposure to the announced measures. In terms of conceptual framework and methods, we build on the literature on central bank communication, which uses language analysis and treats policy announcements as signals that shape expectations (Hansen et al., 2018; Ehrmann and Talmi, 2020). A related strand of work studies the political economy of budget trade-offs (Adolph et al., 2020) and the role of information flows in distinguishing anticipated from unanticipated fiscal policy in budget announcements (Leeper et al., 2013). Balladares and Garcia-Miralles (2025) trace fiscal drag in Spain over time, while in the Irish context (Boyd and McIndoe-Calder, 2025) the focus is on personal income tax over a shorter horizon (2019-2023), using microsimulation modelling with EUROMOD. However, there is little long-run evidence on Irish budget cycles or their political determinants beyond the aggregate analysis in Cousins (2007). In the first part of this paper, we build a new historical dataset of Irish budgets from the mid-1990s onward, coding discretionary changes by broad policy area. We benchmark each years policy package against a counterfactual that mechanically indexes key tax and welfare parameters to price and wage inflation. Deviations from this benchmark provide a transparent measure of discretionary policy priorities. We then relate these deviations to (i) topic-level media salience indices constructed using text analysis of newspaper coverage, (ii) measures of lobbying intensity proxied by pre-budget submissions and public representations by sector, and (iii) electoral timing and macroeconomic conditions. In the second part, we analyse individual-level survey data to examine perceptions of redistribution, fairness, and uncertainty about the future in relation to budget announcements. Microsimulation using EUROMOD allows us to account for heterogeneity in responses, in particular whether individuals or households that are likely to gain or lose from policy changes react differently. We exploit variation in interview timing relative to Budget Day using the Irish sample of the European Social Survey, complemented by Eurobarometer data. Our work provides a systematic assessment of Irish budget priorities and contributes to the debate on whether indexation would reduce fiscal drag and depoliticise distributional choices, or whether discretionary budgets are instead responsive to public preferences.