Welfare losses, preferences for redistribution, and political participation: Evidence from the United Kingdom's age of austerity

Martorano, Bruno; Metzger, Laura; Justino, Patricia Working Paper Welfare losses, preferences for redistribution, and political participation: Evidence from the United Kingdom's age of austerity UNU-MERIT Working Papers, No. 2023-016 Provided in Cooperation with: Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT), United Nations University (UNU) Suggested Citation: Martorano, Bruno; Metzger, Laura; Justino, Patricia (2023) : Welfare losses, preferences for redistribution, and political participation: Evidence from the United Kingdom's age of austerity, UNU-MERIT Working Papers, No. 2023-016, United Nations University (UNU), Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT), Maastricht This Version is available at: https://hdl.handle.net/10419/326866 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-sa/4.0/ #2023-016 Welfare losses, preferences for redistribution, and political participation: Evidence from the United Kingdom’s age of austerity Patricia Justino, Bruno Martorano and Laura Metzger Published 8 May 2023 Maastricht Economic and social Research institute on Innovation and Technology (UNU-MERIT) email: [email protected]u | website: http://www.merit.unu.edu Boschstraat 24, 6211 AX Maastricht, The Netherlands Tel: (31) (43) 388 44 00 UNU-MERIT Working Papers ISSN 1871-9872 Maastricht Economic and social Research Institute on Innovation and Technology UNU-MERIT | Maastricht University UNU-MERIT Working Papers intend to disseminate preliminary results of research carried out at UNU-MERIT to stimulate discussion on the issues raised. Working Paper Welfare losses, preferences for redistribution, and political participation Evidence from the United Kingdom’s age of austerity Patricia Justino,1 Bruno Martorano,2 and Laura Metzger3 May 2023 1 UNU-WIDER, Helsinki, [email protected]; 2 UNU-MERIT and Maastricht University, Maastricht, [email protected]; 3 Radboud University, Nijmegen, [email protected] This study has been prepared within the UNU-WIDER project Institutional legacies of violent conflict. Copyright © UNU 2023 Abstract: This paper studies the effect of austerity on forms of political participation—including voting, appealing for reform, and peaceful protesting—and the role of preferences for redistribution in shaping the relationship between individual exposure to austerity and political participation. The paper focuses on the case of the United Kingdom (UK) where, between 2011 and 2019, wide-ranging austerity policies were introduced to deal with high public debt in the aftermath of the 2007–08 financial crisis. Cuts to government spending on public investment, services, and social protection, especially during the initial fiscal consolidation phase of 2011–15, led to significant welfare losses for the population. We provide evidence from observational microeconomic data and a large-scale online experiment in the UK showing that individual exposure to welfare losses from austerity increases political participation and strengthens preferences for government redistribution. The experimental data suggests that changes in individual preferences for redistribution significantly shape the effect of austerity on political participation. Key words: austerity, political participation, redistribution, expenditure cuts, welfare loss JEL classification: D0, H5, H6, I38 Acknowledgements: This paper is part of a broader UNU-WIDER project on the ‘Institutional Legacies of Violent Conflict’. The authors thank Francesco Iacoella, Alex Hunns, and Isaac Thornton for excellent research assistance, and Sten Ritterfeld for producing the treatment videos. The authors are grateful to Eleonora Nillesen, Andrea Ruggeri, and participants at the seminars at UNU-WIDER (Workshop on Institutional Legacies of Violent Conflict), Wageningen University, UNU-MERIT, and at the MSM Welcome Conference ‘Pathways to Inclusive and Sustainable Development’, University of Maastricht for very useful discussions and comments. This study is published jointly with UNU-MERIT in the respective Working Paper Series. 3 1 Introduction The relationship between welfare losses and political outcomes has received wide attention in the social sciences. A large body of work has studied the effect of economic shocks—mainly due to globalization—on voting outcomes (Margalit 2011; Feigenbaum and Hall 2015), political polarization (Autor et al. 2020), and political preferences (Iacoella et al. 2019; Ahlquist et al. 2020; Caselli et al. 2020; Dippel et al. 2021). Recent studies have also explored the link between austerity policies and political fragmentation in Europe since the financial crisis of 2007–08 (Rüdig and Karyotis 2014; Fetzer 2019; Gabriel et al. 2022; Wiedemann 2022), and across time (Ponticelli and Voth 2020; Alesina et al. 2021; Hübscher et al. 2021). There are two important questions that have received less attention in the literature. First, while most of the existing literature has focused on the effects of austerity on voting outcomes, less attention has been paid to political mobilization. In particular, the literature looks mainly at the consequences of austerity on voting preferences and less at voting turnout which is a key ingredient of democracy. Moreover, it hardly pays attention to other forms of political participation that are important and distinct tools used by citizens to demand change outside election cycles, such as appealing for political reform (e.g., via written or oral appeals to political representatives, participating or organizing petitions and other citizendriven campaigns) and engaging in demonstrations and protests.1 Second, despite a growing interest in the political costs of austerity, few studies focus on causal mechanisms which may determine political participation under austerity. One important, but understudied, mechanism refers to changes in preferences for redistribution. Understanding the role of redistributive preferences in shaping the relationship between austerity and political participation is crucial to recognize and explain what citizens demand from their government in times of crisis and why (Justino and Martorano 2019; Harrison 2021).2 Against this background, this paper answers two research questions. First, we study the effect of austerity on political participation by analysing the political effects of the recent austerity programme in the United Kingdom (UK) following the 2007–08 financial crisis. We focus on three forms of political participation: voting, appealing for reform, and peaceful protesting. Second, we explore the role of individual preferences for redistribution in shaping the relationship between individual exposure to austerity and political participation. The recent experience with austerity policies in the UK offers an important setting to study these questions. Like most European countries, the British government accumulated considerable public debt as a response to dealing with the 2007–08 global financial crisis. In 2008, national public debt was around 37 per cent of the UK’s GDP. By the end of 2012, it reached 70 per cent of the country’s GDP. To lower public debt, the government implemented a set of austerity initiatives. Key policies of the first 2011–15 austerity plan included an increase in consumption taxes from 17.5 to 20 per cent, public sector pay freeze, and £32 billion in welfare spending cuts (Van Reenen 2015). Cuts were particularly steep in 2010 and 2012, when real public investment dropped by 40 per cent (Van Reenen 2015). The British government ended austerity in September 2019, after the second five-year plan (2015– 19). However, in face of a looming recession and a public-debt-to-GDP ratio above 90 per cent, the current government under Rishi Sunak plans to reinstate austerity from 2023, aiming for annual deficit reductions of £60 billion through tax increases and public spending cuts. Media outlets 1 Ponticelli and Voth (2020) use cross-sectional data to study the relationship between austerity and social unrest in several advanced OECD countries between 1919 and 2008. The role of other forms of appeal for political reforms has been less studied. 2 For a discussion in the context of Brexit specifically, see also https://blogs.lse.ac.uk/brexit/2019/10/17/long-readdebunking-myths-on-links-between-austerity-and-brexit/. 4 provocatively referred to these plans as ‘austerity on steroids’;3 and, shortly after the government’s announcement, thousands of citizens in London demonstrated against austerity, demanding general elections to end the Tory government and ‘tax the rich not the poor’.4 This increase in protests has been unprecedented in the recent history of the UK (see Figure 3). Anecdotal accounts link the rise in protests in the UK to austerity and a growing demand for more redistributive policies. However, there is to date limited causal evidence on the effects of austerity on political participation and the role of redistributive preferences in shaping those effects. Even though antiausterity and redistributive calls have been figuring prominently in street protests, many other social changes afoot in the UK (such as anti-migration sentiments, Brexit and, more recently, the COVID-19 pandemic) might confound a direct link between the two phenomena. The breadth and duration of the UK’s post-crisis austerity policies has attracted attention and opposition from different groups. For instance, a press release of the United Nations Human Rights Office in 2019 deplored that fiscal consolidation increased inequality and poverty in the UK and that the British Government broke the social contract by no longer providing a basic social safety net for the population.5 More recently, leading economists and public academic voices stated that austerity has damaged the British economy’s productive capacity and the state’s ability to deal with new crises such as COVID-19 and climate change.6 British citizens as well continue to express grievances against austerity in street protests and grassroot campaigns known as the anti-austerity movement. A growing body of academic literature lends credibility to this debate by documenting the significant costs of austerity to society. Beatty and Fothergill (2016) calculate that average welfare losses from reduced spending on welfare and public services accumulated to £690 per year per working-age adult between 2010 and 2017. Innes and Tetlow (2015) show that local government units, the Local Authority Districts (LADs), had to reduce per-person spending by 23 per cent in real terms between 2010 and 2015 due to cuts in government transfers. Importantly, spending cuts varied much across LADs and hit poorer areas disproportionately hard (Innes and Tetlow 2015; Fetzer 2019), thus distributing the costs of fiscal consolidation unequally across citizens and local administrations. In addition to direct welfare losses, austerity had also knock-on effects on economic growth, which was reduced by 2 per cent between 2010 and 2012 (Van Reenen 2015), crime (Giulietti and McConnell 2022), hate crimes against migrants (Bray et al. 2022), excess mortality (Walsh et al. 2022), and food insecurity (Jenkins et al. 2021).7 3 ‘UK Braces for New Austerity with Little Left for Jeremy Hunt, Rishi Sunak to Cut’. Bloomberg. Available at: https://www.bloomberg.com/news/features/2022-11-16/uk-braces-for-new-austerity-with-little-left-for-jeremyhunt-rishi-sunak-to-cut?leadSource=uverify%20wall (accessed 18 January 2022) 4 ‘“Tax the Rich”: Thousands March in London Anti-Austerity Rally – Video’. The Guardian. Available at: https://www.theguardian.com/uk-news/video/2022/nov/05/tax-the-rich-thousands-march-in-london-antiausterity-rally-video (accessed 18 January 2022). 5 ‘UN Expert Laments UK’s “Doubling Down on Failed Anti-Poor Policies”’. OHCHR Press Release. Available at: https://www.ohchr.org/en/press-releases/2019/05/un-expert-laments-uks-doubling-down-failed-anti-poorpolicies (accessed 18 January 2022). 6 ‘Top Economists Warn the UK Not to repeat Austerity After the Covid-19 Crisis’. The New Statesman. Available at: https://www.newstatesman.com/long-reads/2020/05/top-economists-warn-uk-not-repeat-austerity-after-covid-19crisis (accessed 18 January 2022). 7 While these studies illustrate reductions in living standards, especially for vulnerable population groups, empirical evidence on aggregate distributional effects of austerity in the UK is scarce. Great Britain’s Equality and Human Rights Commission (EHCR) reported in 2018 that households in the two bottom deciles had lost approximately 10 per cent of their net income because of austerity (the most vulnerable groups experienced losses above 20 per cent), whilst higher income households were much less affected (EHCR 2018). Regarding the experiences of other countries, a working paper by Ball et al. (2013) shows for 17 OECD countries over the period 1978–2009 that fiscal consolidation 5 In this paper, we investigate whether exposure to this period of austerity affected the political participation of UK citizens and whether changes in redistributive preferences may explain those effects. To do so, we first make use of observational panel data to examine how citizens respond to austerity at the level of Local Authority Districts (LADs). LADs are local government bodies and the final link in the chain responsible for implementing austerity policies. Data on financial welfare losses caused by austerity per LAD (our main explanatory variable) comes from Beatty and Fothergill (2016). We match these welfare losses to political participation (voting, protesting, and appealing) using data from the UK Parliament House of Commons Library, and Google’s Global Database for Events, Language and Tone (GDELT).8 A challenge in researching the relationship between austerity, political participation, and individual preferences for redistribution with observational data is the lack of a counterfactual, i.e. we cannot assess what would have happened in the absence of austerity. This limits opportunities to investigate the causal impact of austerity on political participation and its determinants. To generate such causal evidence, we conduct a large representative online experiment that explores the impact of individual exposure to austerity on political participation and preferences for redistribution in the UK. We leverage the experiment to scrutinize findings from the observational data analysis and study the causal impact of exposure to austerity on individual political participation. The experiment mimics exposure to austerity by showing participants a professionally produced, 3minute video on UK austerity policies and their socioeconomic implications. Using simple language and visuals, the video informs participants how UK austerity came about and what core policies it entailed. It then illustrates and summarizes evidence on the costs that austerity has inflicted on British society, especially on citizens with low-income levels. The control group watches an unrelated video about weather forecast technologies in Europe. Since we cannot simulate protests or elections in the experiment, we generate the data for individual political participation from survey questions on voting preferences and stated willingness to join a peaceful protest against austerity. In addition, we elicit participants’ willingness to sign a real petition for or against austerity—which we physically sent to a Member of Parliament—to complement the selfreported data. In the observational data and the experiment, we find evidence that exposure to austerity mobilizes democratic participation through voting in elections and appealing for political reforms with the government but has no effect on protests. We then proceed to study the role of redistributive preferences in explaining these results.9 There are important reasons why redistributive preferences may mediate the effect of austerity exposure to political participation. First, austerity redistributes income between groups. Ripple effects of this redistribution, like rising inequality and, generally, direct, or indirect exposure to hardship, can change people’s social preferences and the kind of government policies they are willing to raised inequality, decreased wage income shares, and increased long-term unemployment. Their analysis further suggests that spending cuts have had, on average, larger distributional effects than tax increases. 8 The GDELT Project is a ‘Big Data’ source employing an algorithm to process and organize information on events worldwide, including protests and demands for political reform, from digitized newspapers and news agencies, as well as from web-based news aggregators. 9 Preferences for redistribution are of course only one of several relevant channels through which austerity may affect political participation. Other variables that may serve as mediators include trust in political institutions (Biten et al. 2022; Bosco and Verney 2012; Hernández and Kriesi 2016; Hübscher et al. 2021; Marsh and Mikhaylov 2014; Talving 2017) and perceptions of capacity to affect the political process (Harrison 2022). We focus on redistributive preferences in this paper, as this mechanism has remained unexplored in the literature but appears to have a particularly important role in the case of the UK. 6 support.10 There is a large but inconclusive theoretical and empirical literature linking inequality and preferences for redistribution (see Ciani et al. 2021 for an overview and summary). By contrast, there is much less research on the impact of preferences for redistribution and political participation (one exception is Justino and Martorano 2019), specifically in the context of austerity. Second, survey data from the UK shows that post-2010 austerity coincides with increased support for government redistribution. The British Election Study (BES) reveals that preferences for reducing inequality rose markedly after 2010 compared to the previous decade (see Figure A1, in Appendix A). Similarly, the British Social Attitudes Survey11 shows that, by 2017, 48 per cent of citizens supported more public spending through higher taxation and 42 per cent agreed that the government should redistribute income from wealthier to poorer citizens (Park et al. 2012; Clery et al. 2017; Phillips et al. 2018). This compares to 31 per cent of citizens supporting more public spending through higher taxation in 2010, and 37 per cent agreeing that the government should redistribute income from the wealthier to the less affluent (Park et al. 2012). Before 2010, support for government spending and redistribution was falling. Third, a study by Brown et al. (2022) shows that local spending reductions between 2011 and 2015 strengthened support for government redistribution from high-income households, especially for services that these households might benefit from themselves. The experiment we conduct establishes that exposure to austerity causally increases individual preferences for redistribution in the form of government spending and taxing higher incomes and that redistributive preferences explain between 8 and 11 per cent of the effect of austerity exposure on political participation. Our research contributes to the growing literature focusing on the political consequences of austerity in three main ways (Rüdig and Karyotis 2014; Alesina et al. 2021; Fetzer 2019; Ponticelli and Voth 2020; Galofré-Vilà et al. 2021; Hübscher et al. 2021; Gabriel et al. 2022; Wiedemann 2022). First, we add to this literature by showing that welfare losses due to austerity encourage political participation, which is in line with the predictions of pioneering studies of Runciman (1966) and Gurr (1970). Specifically, we provide directly comparable evidence of the impact of austerity on the full spectrum of political participation. Exposure to austerity mobilizes democratic participation not only through voting in elections but also via appealing for political reforms with the government. Second, we build a bridge between this literature and microeconomic research on the social and economic costs of austerity through the focus on redistributive preferences. Understanding why (and the way through which) welfare losses affect people’s decision to engage politically is very important for ensuring the proper functioning of democracy (Fetzer 2019). Finally, we add to a small set of new studies that explore the causal impact of post-financial crisis austerity on socioeconomic outcomes as well as political participation and democracy (e.g., Fetzer 2019; Bray et al. 2022; Gabriel et al. 2022; Hübscher et al. 2022; Wiedemann 2022). Exploring causal determinants of political participation can help us understand what citizens demand from their government in times of crisis and why they demand it, thus providing important insights into how economic crises might affect democratic participation. 2 Observational data and key trends The unit of analysis in this first part of the study is the Local Authority District (LAD). LADs are the geographic area for which local government bodies are responsible and comprise district 10 This happened, de facto, during austerity. For instance, trading off the political costs of different measures, the British Government shielded pensions during austerity and cut social housing benefits (two thirds of people affected were disabled) and spending on education instead. 11 See https://www.bsa.natcen.ac.uk. 13 Both information treatments allow us to test the hypothesis that exposure to austerity has a causal effect on political participation. However, the information in treatment group 1 is not personalized and refers to austerity in the UK in the aggregate. This setting considers that not all citizens are equally affected by austerity: some may experience it from a distance. Treatment group 2 offers a more personalized experience of austerity and, hence, considers that some citizens may have been impacted by welfare cuts themselves or have been directly observing hardship in their communities. Designing the information treatments in this way enables us to differentiate between the effect of ‘merely’ providing information about austerity and its costs to society (treatment group 1), and the effect of creating a more personal and relatable experience of being exposed to austerity (treatment group 2) that is closer to lived realities. The alternative framing, in this case, should modify the criteria according to which a policy is perceived and judged (Scheufele 2000; Stantcheva 2022) and, eventually, affect the desire to take political action. Appendix F shows the scripts of the treatment videos.23 To keep conditions constant across experimental groups, the control group watched an unrelated video about European weather forecast technology. This ensures comparable initial conditions across all groups since delivering information through a video may affect individual characteristics such as reducing the risk of uncertainty, having a priming effect, increasing attention, and generating emotional responses (Haaland et al. 2023). All videos are similar in length: 3.06 minutes for the control group, 2:26 minutes for treatment group 1, and 3:25 minutes for treatment group 2. To ensure that participants watch the video (are treated), we remove any technical features that enabled forwarding, playing back, or skipping the video. A few seconds after the video ends, the ‘next’ button appears on the survey page and participants can proceed with the survey. After watching the videos, participants answer questions about political participation. These questions generate our first set of dependent variables that tightly align with the dependent variables of the observational study (voting, appealing for political reform, and peaceful protesting). In contrast to the observational data, where we cannot observe whether citizens vote, protest, or appeal for reform to oppose austerity specifically, we are now able to ask about respondents’ political participation due to austerity policies. The first two questions elicit stated preferences; the third question asks participants to reveal their preferences and sign a petition which is forwarded to two real members of Parliament.24 1. Voting. ‘How likely would you be to vote for a candidate for public office because of their position on austerity?’ – Range: 1 (Definitely would not) to 4 (Definitively would) 2. Protesting. ‘If a peaceful protest against austerity policies was being held this week close to where you live, would you participate?’ Range: 1 (Definitely will not) to 5 (Definitively will) 3. Appealing for political reform. ‘You can also take action by signing a petition. We will submit each petition to two Members of Parliament who have either spoken for or against austerity policies in parliament. We will not tell them your name, just how many 23 The videos are available here: https://youtube.com/playlist?list=PLGO1JriFGwM6OykTgF8DGmZTq46lPLoAI. 24 We identified two real members of the parliament (MPs) who argued against austerity or for austerity policies in publicly available speeches held in parliament in the recent past. We contacted them via email briefly explaining the content of our study. We highlighted that participants had the opportunity to express their desire to demand from UK members of parliament to take a stand against austerity or for austerity policies. We reported to them the number of UK citizens who participated in our study and expressed such a desire. On behalf of the participants, we explicitly requested to the MP against austerity to take a stand against the erosion of the welfare state. We requested to the MP for austerity to stand against excessive budget deficits and public sector spending in relevant government debates. 14 people in our study support either of the two petitions below. We will send you proof of submitting the petition to the corresponding Member of Parliament in the next few weeks.’ Respondents could choose one of three answers: - I want to sign the following petition AGAINST austerity policies. ‘Austerity policies affect all segments of the population and particularly hurt the most vulnerable groups. Spending cuts should be stopped. I am against austerity policies.’ - I want to sign the following petition FOR austerity policies. ‘Cuts are necessary to ensure economic sustainability. Benefits should be reserved for people who really need them such as old people and children. I support austerity policies.’ - I do not want to sign either of these petitions. Next, we gather post-treatment data on individual preferences for redistribution, our second set of dependent variables. We ask the following questions:25 ● Redistribution. ‘The government should implement policies to reduce differences in income levels between the rich and the poor.’ Range: 1 (strongly disagree) to 5 (strongly agree). ● Taxation and public spending. ‘If the UK’s national budget deficit is too high, the government should raise taxes on the rich instead of cutting benefits for the poor.’ Range: 1 (strongly disagree) to 5 (strongly agree). ● Government spending. ‘Suppose you can decide over total UK government spending next year. How do you want to divide the budget in percentages between the following categories? (The total must sum up to 100%).’ Categories: health, education, defence and national security, welfare spending (e.g. income support or child benefits), social security, and public infrastructure. We use these questions to investigate to what extent preferences for redistribution shape the relationship between exposure to austerity and political participation. First, we test whether exposure to austerity directly changes individual preferences for redistribution. Second, we run a mediation analysis to test whether redistribution preferences are a channel through which austerity affects political participation. 3.3 Data quality and descriptive statistics In total, 1,494 people living in the UK participated in our survey experiment. We took several measures to ensure data quality. First, we excluded one participant who joined from outside of the UK according to GPS data. Second, we used an attention check implemented immediately before 25 Questions on preferences for redistribution are taken from the relevant literature (e.g. Alesina and La Ferrara 2000) and adjusted for the purpose of our survey. The British Election Study (BES) post-electoral survey reports a similar question: ‘The government should take measures to reduce differences in income levels’. Regarding taxation, we took inspiration from the study of Kuziemko et al. (2015) who asked participants the following question: ‘As you may know, there have been proposals recently to decrease the federal deficit by raising income taxes on millionaires. Do you think income taxes on millionaires should be increased, stay the same or decreased?’. Questions about spending are similar to Alesina et al. (2018) but adapted to the UK context. 15 the treatment to test if participants devoted their full attention to the survey,26 and informed respondents that their answer would not affect payment. Two participants reported that they did not pay attention to the survey. Although answering ‘no’ to this question might have happened by mistake, we excluded these two observations from the sample. We excluded one more participant who did not give permission to use their data. Fourteen additional participants were dropped from the analysis because they have missing information about one of the control variables. In total, we excluded 18 observations from the full sample, leaving us with 1,476 participants (Table 3). A balance test shows that excluding these observations does not affect the comparability of the experimental groups. Results are reported in Table G1 in Appendix G. We observe a low spread in values for the average time taken to complete the survey. The mean (13 minutes) and median (12 minutes) are similar and 98 per cent of participants spent a reasonable time completing the survey (within 3 standard deviations of the mean). Excluding an additional 27 participants who took longer to complete the survey does not affect our main results (Table G2 in Appendix G). 3.4 Empirical strategy We estimate the impact of Exposure to Austerity (treatment 1) and Exposure to Austerity Deprived (treatment 2) on our outcomes of interest using the following OLS regression: 𝑦 𝛼 𝛼𝑇  𝛼𝑋𝑢  (2) where i defines the individual and j the treatment status. Y indicates the outcomes of interest— political participation (voting, protesting, or signing a petition against austerity) or preferences for redistribution. The treatment dummy is represented by T and equals 1 for any treatment group and 0 for the control group. X represents the control variables, all measured prior to administering the treatment. These include demographic characteristics (sex, age bracket, ethnic background, marriage status, and whether they have lived in the UK for most of their life), socioeconomic characteristics (whether the respondent has a university or postgraduate degree, is in paid employment (full-time or part-time), and their income class),27 levels of trust in the government, with values ranging from 1 (all the time) to 5 (never), and political ideology, proxied by a dummy variable equal to 1 if the respondent considers themselves as ‘left wing’ or ‘left center’ and 0 for the other political affiliation. Descriptive statistics for all control variables are reported in Table H1 in Appendix H. Regression results without controls and results without trust or political preferences are reported in Appendix I. 3.5 Main results Table 4 presents the analysis of the impact of exposure to austerity on political participation, and Table 5 presents the analysis of the impact of exposure to austerity on individual preferences for redistribution. In both cases, we first estimate the effect of the pooled treatment sample (Exposure to Austerity + Exposure to Austerity Deprived) and then break it down by treatment group. Table 6 presents the results of the mediation analysis which tests whether preferences for redistribution could be a channel through which austerity affects political participation. 26 This strategy was also intended to increase participants’ attention just before the informational treatment. 27 Twelve classes are considered: less than 15,000; 15,000–24,999; 25,000–34,999; 35,000–44,999; 45,000–54,999; 55,000–64,999; 65,000–74,999; 75,000–84,999; 85,000–94,999; 95,000–104,999; 105,000–144,999; more than 145,000. 16 The results in Table 4 (Panel A) show that exposure to austerity increases individual willingness to vote and appeal for political change (sign a petition), but not the willingness to join a peaceful protest. The results confirm our findings from the observational data. One explanation for the null effect on protests could be that this form of political participation requires greater individual effort and commitment than voting or signing a petition. The experimental treatment might not have successfully (re)-created the right conditions to invite such a response. The pandemic and risk of contracting COVID-19 also still dominated public life when we implemented the experiment and might have affected answers about joining a public event. However, this result replicates the observational analysis, which suggested that exposure to austerity affected all forms of political participation except for protesting. Disaggregating the results by treatment (Table 4, Panel B) shows that providing additional personalized information on how vulnerable population groups experience austerity magnifies the treatment impact. However, we do not find a statistically significant difference between the two treatments. The results in Table 5 (Panel A) show that exposure to austerity increases individual preferences for redistribution across all three measures (government policies that reduces income differences between rich and poor, taxation and public spending, and government spending for welfare benefits and social security). Three results emerge when we separate the effects by treatment. First, providing additional personalized information about hardship for vulnerable citizens (treatment group 2) significantly increases support for government policies to reduce income differences between rich and poor (Column 1). Second, personalized information about hardship significantly drives support for taxing richer citizens rather than cutting social protection measures to reduce the budget deficit (Column 4). Third, personalized information about hardship has a larger positive effect on individual support for allocating more resources to welfare and social security spending than the aggregate information presented in treatment group 1, although this difference is not significant. We also tested whether response to treatment differed conditional on pre-treatment measures of political ideology and trust in the government using an interaction term and do not find significant differences with regards to these characteristics (see Appendix K). 3.6 Mediation analysis In this section, we explore whether changes in individual preferences for redistribution are a possible mechanism through which exposure to austerity (Table 5) increases political participation. We focus on voting and appealing for political reform and exclude protesting since we do not find a significant treatment effect on this variable, and focus on the ‘Government should reduce income inequality’ as our preferred measure of individual preferences for redistribution. This variable refers specifically to government redistribution and applies broadly rather than making references to specific policies. However, we also report the results for the other two measures of preferences for redistribution in the Appendix J. As discussed earlier, preferences for redistribution are an important understudied channel through which exposure to austerity may affect political participation. First, austerity policies redistribute income between socioeconomic groups, which has been shown to change people’s social preferences and the kind of government policies they are willing to support. Second, a recent study by Brown et al. (2022) on the UK suggests that austerity can indeed change individual preferences for redistribution. Providing reliable evidence on preferences for redistribution as a mediator between austerity and political participation is challenging with observational data. The entire population has been exposed to austerity and there is no exogenous variation in this variable. 17 Identifying the share of political participation explained by austerity through a change in preferences for redistribution is not trivial with experimental data either. For this to work, preferences for redistribution have to be unaffected by both the treatment being administered and any other pre-intervention factors (Imai 2011). Instead of following the usual approach to mediation analysis (see Imai 2011 for a discussion), we follow an empirical strategy implemented by Heller et al. (2017), taking advantage of the fact that the mediator is independent of the treatment in the control group. First, we estimate the impact of the treatment (𝑇), Exposure to Austerity, on the candidate mechanism (M), preferences for redistribution. Formally: 𝑀  𝛽 𝛽𝑇  𝛽𝑋𝑢  (3) where 𝛽 is the coefficient of interest which captures the effect of the treatment (exposure to austerity) on the mediator (preferences for redistribution). The analysis of the effect of Exposure to Austerity on preferences for distribution shows that treated individuals are more likely to support government policies to reduce income differences between rich and poor (i.e. income inequality). The analysis of the effect of Exposure to Austerity on preferences for distribution also shows that treated individuals are more likely to support increasing welfare and social security spending and raising taxes on richer citizens instead of cutting social protection to reduce a budget deficit (Table J1 in Appendix J). Exposure to Austerity Deprived significantly drives support for taxing richer citizens rather than cutting social protection measures to reduce the budget deficit (Table J1-Panel B in Appendix J). Last, Exposure to Austerity Deprived has a somewhat larger positive effect on preferences for allocating more resources to welfare and social security spending than Exposure to Austerity, although this difference is not significant (Table J1-Panel B in Appendix J). Second, we estimate the impact of the mediator (M) on the outcome of interest (y), i.e. political participation. Formally: 𝑦  𝛾𝛾 𝑇 𝛾 𝑀 𝛾 𝑋𝑢 (4) Multiplying 𝛽∗ 𝛾 yields the indirect effect of the treatment 𝑇 (exposure to austerity) on y (political participation) through M (preferences for redistribution). This approach is problematic, however, because two assumptions need to be satisfied for it to work (Imai et al. 2011). The first assumption is that the treatment is independent of potential mediators and outcomes. In a randomized study—such as ours—this assumption is satisfied. The second assumption is unlikely to hold, even in an experimental setting, because it requires that the mediator is statistically independent of the treatment (and other pre-intervention characteristics). Heller et al. (2017) propose a strategy, which partially mitigates this problem, and which we apply to our context, whereby the potential mediator must be independent of the treatment in the control group. In this approach, we re-estimate Equation (4) using data from the control group: 𝑦  𝛿𝛿 𝑀 𝛿 𝑋𝑢  for all (i) with 𝑇 0. (5) The indirect effect of the treatment 𝑇 (exposure to austerity) on y (political participation) through M (preferences for redistribution) is now given by 𝛽* 𝛿. We obtain the share of political participation explained by austerity through a change in preferences for redistribution using 𝛽* 𝛿/ 𝛼. These estimations are shown in Table 6. Results suggest that preferences for redistribution are a relevant channel through which austerity impacts political participation. Column 3 shows that preferences for redistribution (Government should reduce inequality) mediate around 8 per cent of the effect of Exposure to Austerity on voting and around 11 per cent of its effect on petitioning for political reform. In addition, the effect of Exposure to Austerity on political participation is mediated by the increasing support for ‘raising taxes on richer citizens’ (Table J2 in Appendix J). 18 4 Discussion and conclusion This study provides evidence from observational microeconomic data and a large-scale online experiment in the UK that exposure to austerity increases democratic participation through voting in elections and appealing to the government for political reforms, but not protesting. The results also show that exposure to austerity increases individual preferences for government redistribution via taxing higher incomes and spending more on welfare and social security. A mediation analysis based on the experimental data suggests that preferences for redistribution are a relevant determinant of political participation under austerity. These findings have important policy implications. There has been a growing academic and public debate about the political costs of austerity in the UK and elsewhere, including its effect on democracy. Emerging studies show an association between welfare cuts, the erosion of social cohesion, and political fragmentation (see Guriev and Papaioannou 2022). Our results indicate that exposure to austerity increased political participation motivated by a demand for redistribution. Possibly UKIP effectively tapped into this sentiment, especially in areas where citizens felt left behind by the austerity policies of the incumbent government. However, voters’ recognition of false claims about economic prosperity and the redistribution of resources earmarked for the EU back into the hands of British citizens after Brexit eventually eroded support for UKIP. In that spirit, the quick downfall of Prime Minister Liz Truss after 42 days in office might be another example of the inability of politics to understand citizens’ preferences for redistribution. Her proposed agenda of cutting taxes on the rich did not appeal to ordinary citizens, including Brexit voters.28 Our research findings provide new insights into the relationship between citizens and democratic governments during long-term crises. Current policy debates on the COVID-19 pandemic, the war against Ukraine, and the climate crisis across the population partly resemble the austerity debate, with some problems being of similar nature, especially growing calls for welfare redistribution to ensure the most vulnerable groups are not adversely affected further by these combined crises. Crises change preferences which in turn shape democratic participations. Governments may ignore such demands at their own peril. References Aguilar, E. E. and A. C. Pacek. 2000. “Macroeconomic Conditions, Voter Turnout, and the Working Class/Economically Disadvantaged Party Vote in Developing Countries.” Comparative Political Studies 33 (8): 995–1017. https://doi.org/10.1177/0010414000033008001 Alaimo, L. S. and L. M. Solivetti. 2019. “Territorial Determinants of the Brexit Vote.” Social Indicators Research 144 (2): 647-667. https://doi.org/10.1007/s11205-019-02066-3 Alesina, A. and E. La Ferrara. 2000. “Participation in Heterogeneous Communities.” The Quarterly Journal of Economics, Vol. 115(3), 847–904. https://doi.org/10.1162/003355300554935 Alesina, A., S. Stantcheva and E. Teso. 2018. “Intergenerational Mobility and Preferences for Redistribution.” American Economic Review 108 (2): 521-54. https://doi.org/10.1257/aer.20162015 28 See https://www.bbc.com/news/uk-63838387 and https://www.express.co.uk/news/politics/1685171/brexitnews-liz-truss-conservative-labour-party-labour-immigration-businesses-economy-vn. 19 Alesina, A., C. Favero and F. Giavazzi. 2019. “Effects of Austerity: Expenditureand Tax-Based Approaches.” Journal of Economic Perspectives 33 (2): 141–62. https://doi.org/10.1257/jep.33.2.141 Alesina, A., D. Furceri, G. Ciminelli and G. Saponaro. 2021. “Austerity and Elections.” Technical report, International Monetary Fund, Washington D. C. https://doi.org/10.5089/9781513573724.001 Autor D., Dorn, D., Hanson, G. and Kaveh, M. 2020. “Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure.” American Economic Review 110 (10): 3139-3183. https://doi.org/10.1257/aer.20170011 Aytaç, S E., E. G. Rau and S. Stokes. 2020. “Beyond Opportunity Costs: Campaign Messages, Anger and Turnout among the Unemployed.” British Journal of Political Science 50 (4): 1325-1339. https://doi.org/10.1017/S0007123418000248 Ball, L.M., D. Furceri, M.D. Leigh and M.P. Loungani. 2013. “The Distributional Effects of Fiscal Consolidation.” International Monetary Fund, Washington D. C. https://doi.org/10.5089/9781475551945.001 Beatty, C. and S. Fothergill. 2013. “Hitting the Poorest Places Hardest: The Local and Regional Impact of Welfare Reform.” Technical Report, Sheffield Hallam University. https://doi.org/10.7190/cresr.2017.6378897426 Beatty, C. and S. Fothergill. 2016. “The Uneven Impact of Welfare Reform the Financial Losses to Places and People.” Technical Report, Sheffield Hallam University. https://doi.org/10.7190/cresr.2017.5563239352 Biten, M., K.T. and W. van der Brug. 2022. “How Does Fiscal Austerity Affect Trust in the European Union? Analyzing the Role of Responsibility Attribution.” Journal of European Public Policy. https://doi.org/10.1080/13501763.2022.2060282 Bobo, L. and F.D. Gilliam. 1990. “Race, Sociopolitical Participation, and Black Empowerment.” American Political Science Review 84 (2): 377-393. https://doi.org/10.2307/1963525 Bray, K., B.N. and J. Wildman. 2022. “Austerity, Welfare Cuts and Hate Crime: Evidence from the UK’s Age of Austerity.” Journal of Urban Economics 103439. https://doi.org/10.1016/j.jue.2022.103439 Brown, S., A. Kontonikas, A. Montagnoli, M. Moro, D. Papoutsaki and W. Sas. 2021. “Who Cares? Attitudes Towards Redistribution and Fiscal Austerity.” CESifo Working Paper, No. 9393, Center for Economic Studies and ifo Institute (CESifo), Munich. https://doi.org/10.2139/ssrn.3959447 Burden, B. C. 2009. “The Dynamic Effects of Education on Voter Turnout.” Electoral Studies 28 (4): 540– 549. https://doi.org/10.1016/j.electstud.2009.05.027 Bömmel, N. and G. Heineck. 2020. “Revisiting the Causal Effect of Education on Political Participation and Interest.” IZA Discussion Paper Series, IZA DP No.13954, December 2020. https://doi.org/10.2139/ssrn.3751847 Chevalier, A. and O. Doyle. 2012. “Schooling and Voter Turnout: Is there an American Exception?” IZA Discussion Paper Series, IZA DP No. 6539, May 2012. https://doi.org/10.2139/ssrn.2056729 Ciani, E., L. Fréget and T. Manfredi. 2021. “Learning About Inequality and Demand for Redistribution: A Meta-Analysis of In-Survey Informational Experiments.” OECD Papers on Well-being and Inequalities, No. 02, OECD Publishing, Paris. https://doi.org/10.1787/8876ec48-en Clark, D. H., and P. M. Regan. 2018. “Mass Mobilization Protest Data.” The Mass Mobilization Project. URL: https://www.binghamton.edu/massmobilization/about.html. Clery, E., J. Curtice and R. Harding. 2017. British Social Attitudes: The 34th Report, London: The National Centre for Social Research, Dalton, R. J. 2017. The Participation Gap: Social Status and Political Inequality. Oxford: Oxford University Press. https://doi.org/10.1093/oso/9780198733607.001.0001 20 Delli Caprini, M. X. and S. Keeter. 1996. What Americans Know About Politics and Why It Matters (1st ed.). New Haven, London: Yale University Press. Denny, K. and O. Doyle. 2008. “Political Interest, Cognitive Ability and Personality: Determinants of Voter Turnout in Britain.” British Journal of Political Science 38 (2): 291-310. https://doi.org/10.1017/S000712340800015X Emmenegger, P., P. Marx and D. Schraff. 2016. “Off to a Bad Start: Unemployment and Political Interest during Early Adulthood.” The Journal of Politics 79 (1): 315–28. https://doi.org/10.1086/688226 Erikson, R. S. 2015. “Income Inequality and Policy Responsiveness.” Annual Review of Political Science 18: 11-29. https://doi.org/10.1146/annurev-polisci-020614-094706 Fetzer, T. 2019. “Did Austerity Cause Brexit?” American Economic Review 109 (11): 3849–86. https://doi.org/10.1257/aer.20181164 Fialho, F.M. 2022. “Race and Non-Electoral Political Participation in Brazil, South Africa, and the United States.” Journal of Race, Ethnicity, and Politics 7 (2): 262-293. https://doi.org/10.1017/rep.2021.29 Fraga, B.L. 2018. The Turnout Gap: Race, Ethnicity, and Political Inequality in a Diversifying America. Cambridge M. A.: Cambridge University Press. https://doi.org/10.1017/9781108566483 Funk, P. 2016. “How Accurate Are Surveyed Preferences for Public Policies? Evidence from a Unique Institutional Setup.” Review of Economics and Statistics 98 (3): 442-454. https://doi.org/10.1162/REST_a_00585 Gabriel, R.D., M. Klein and A. S. Pessoa. 2022. “The Political Costs of Austerity.” Sveriges Riksbank Working Paper Series, No. 418, November 2022. Gallego, A. 2015. Unequal Political Participation Worldwide. Cambridge, M. A.: Cambridge University Press. https://doi.org/10.1017/CBO9781139151726 Galofré-Vilà, G., C. M. Meissner, M. McKee, and D. Stuckler (2021). Austerity and the rise of the nazi party. The Journal of Economic History 81(1), 81–113. https://doi.org/10.1017/S0022050720000601 Giulietti, C. and B. McConnell. 2022. “Kicking You When You’re Already Down: The Multipronged Impact of Austerity on Crime.” The ESRC Centre for Population Change Working Paper Series, Working Paper 104, November 2022 Goldberg. M.H., S. van der Linden, M.T. Ballew, S.A. Rosenthal, A. Gustafson and A. Leiserowitz. 2019. “The Experience of Consensus: Video as an Effective Medium to Communicate Scientific Agreement on Climate Change.” Science Communication 41 (5): 659-673. https://doi.org/10.1177/1075547019874361 Grasso, M.T. and M. Giugni. 2016. “Protest Participation and Economic Crisis: The Conditioning Role of Political Opportunities.” European Journal of Political Research 55 (4): 663-680. https://doi.org/10.1111/1475-6765.12153 Guriev, S. and E. Papaioannou. 2022. “The Political Economy of Populism.” Journal of Economic Literature 60 (3): 753-832. https://doi.org/10.1257/jel.20201595 Gurr, T.R. 1970. Why Men Rebel. Princeton, NJ: Princeton University Press. Haaland, I., C. Roth and J. Wohlfart. 2023. “Designing Information Provision Experiments.” Journal of Economic Literature, 61(1): 4-40. https://doi.org/10.1257/jel.20211658 Harrison, K. 2020. “Can’t, Won’t and What’s the Point? A Theory of the UK Public’s Muted Response to Austerity”, Representation 57 (2), 159-174. https://doi.org/10.1080/00344893.2020.1728367 Hauser, S. M. 2000. “Education, Ability, and Civic Engagement in the Contemporary United States.” Social Science Research 29 (4): 556-582. https://doi.org/10.1006/ssre.2000.0681 Heller, S.B., A. K. Shah, J. Guryan, J. Ludwig, S. Mullainathan and H.A. Pollack. 2017. “Thinking, Fast and Slow? Some Field Experiments to Reduce Crime and Dropout in Chicago.” Quarterly Journal of Economics 132 (1): 1-54. https://doi.org/10.1093/qje/qjw033 21 Howell, S.E. and D. Fagan. 1988. “Race and Trust in Government: Testing the Political Reality Model.” Public Opinion Quarterly 52 (3): 343-350. https://doi.org/10.1086/269111 Hübscher, E., T. Sattler and M. Wagner. 2021. “Voter Responses to Fiscal Austerity.” British Journal of Political Science 51 (4): 1751-1760. https://doi.org/10.1017/S0007123420000320 Imai, Kosuke, L. Keele, D. Tingley and T. Yamamoto. 2011. “Unpacking the Black Box of Causality: Learning about Causal Mechanisms from Experimental and Observational Studies.” American Political Science Review 105 (4): 765-789. https://doi.org/10.1017/S0003055411000414 Innes, D. and G. Tetlow. 2015. “Delivering Fiscal Squeeze by Cutting Local Government Spending.” Fiscal Studies 36 (3): 303-325. https://doi.org/10.1111/j.1475-5890.2015.12056 Jenkins, R.H., S. Aliabadi, E.P. Vamos, D. Taylor-Robinson, S. Wickham, C. Millett, C. and A.A. Laverty. 2021. “The Relationship Between Austerity and Food Insecurity in the UK: A Systematic Review.” eClinicalMedicine 33: 100781. https://doi.org/10.1016/j.eclinm.2021.100781 Jungkunz, S. and P. Marx. 2021. “Income Changes Do Not Influence Political Participation: Evidence from Comparative Panel Data.” IZA Discussion Paper Series, IZA DP No. 14198, March 2021. https://doi.org/10.2139/ssrn.3808461 Justino, P. and B. Martorano. 2019. “Redistributive Preferences and Protests in Latin America.” Journal of Conflict Resolution 63 (9): 2128-2154. https://doi.org/10.1177/0022002719827370 Kurer, T., S. Häusermann, B. Wüest and M. Enggist. 2019. “Economic Grievances and Political Protest.” European Journal of Political Research 58 (3): 866-892. https://doi.org/10.1111/14756765.12318 Kuziemko, I., M. I. Norton, E. Saez and S. Stantcheva. 2015. “How Elastic Are Preferences for Redistribution? Evidence from Randomized Survey Experiments.” American Economic Review 105 (4): 1478-1508. https://doi.org/10.1257/aer.20130360 Lawless, J. L. and R. L. Fox. 2001. “Political Participation of the Urban Poor.” Social Problems 48 (3): 362385. https://doi.org/10.1525/sp.2001.48.3.362 Leetaru, K. and P.A. Schrodt. 2013. “GDELT: Global Data on Events, Location, and Tone, 1979–2012.” ISA annual convention 2 (4): 1-49. Margalit, Y. 2019. “Political Responses to Economic Shocks.” Annual Review of Political Science 22 (1): 277-295. https://doi.org/10.1146/annurev-polisci-050517-110713 Marx, P. and C. Nguyen. 2018. “Anti-Elite Parties and Political Inequality: How Challenges to the Political Mainstream Reduce Income Gaps in Internal Efficacy.” European Journal of Political Research 57 (4): 919-940. https://doi.org/10.1111/1475-6765.12258 Marx, P. 2020. “Anti-Elite Politics and Emotional Reactions to Socio-Economic Problems: Experimental Evidence on ‘Pocketbook Anger’ from France, Germany, and the United States.” British Journal of Sociology 71 (4): 608-624. https://doi.org/10.1111/1468-4446.12750 Pacheco, J.S. and E. Plutzer. 2008. “Political Participation and Cumulative Disadvantage: The Impact of Economic and Social Hardship on Young Citizens.” Journal of Social Issues 64 (3): 571-593. https://doi.org/10.1111/j.1540-4560.2008.00578.x Park, A., E. Clery, J. Curtice, M. Phillips and D. Utting. 2012. British Social Attitudes: The 29th Report, London: The National Centre for Social Research. https://doi.org/10.4135/9781446268292 Peer E, D. Rothschild, A. Gordon, Z. Evernden and E. Damer. 2022. “Data Quality of Platforms and Panels for Online Behavioral Research.” Behavioral Research Methods 54 (4): 1643-1662. https://doi.org/10.3758/s13428-021-01694-3 Phillips, D., J. Curtice, M. Phillips and J. Perry. 2018. British Social Attitudes: The 35th Report, London: The National Centre for Social Research. Ponticelli, J. and H.-J. Voth. 2020. “Austerity and Anarchy: Budget Cuts and Social Unrest in Europe, 1919– 2008.” Journal of Comparative Economics 48: 1-19. https://doi.org/10.1016/j.jce.2019.09.007 22 Pontusson, J. and D. Rueda. 2010. “The Politics of Inequality: Voter Mobilization and Left Parties in Advanced Industrial States.” Comparative Political Studies 43 (6): 675–05. https://doi.org/10.1177/0010414009358672 Rosenstone, S. J. 1982. “Economic Adversity and Voter Turnout.” American Journal of Political Science 26 (1): 25-46. https://doi.org/10.2307/2110837 Rosenstone, S. J. and J.M. Hansen. 1993. Mobilization, Participation, and Democracy in America. New York: Macmillan. Runciman, W.G. 1966. Relative Deprivation and Social Justice. Berkeley, CA: University of California Press. Rüdig, W. and G. Karyotis. 2014. “Who Protests in Greece? Mass Opposition to Austerity.” British Journal of Political Science 44 (3): 487513. https://doi.org/10.1017/S0007123413000112 Schlozman, K. L., S. Verba and H. Brady. 2012. The Unheavenly Chorus: Unequal Political Voice and the Broken Promise of American Democracy. Princeton: Princeton University Press. https://doi.org/10.23943/princeton/9780691154848.001.0001 Solt, F. 2008. “Economic Inequality and Democratic Political Engagement.” American Journal of Political Science 52 (1): 48-60. https://doi.org/10.1111/j.1540-5907.2007.00298.x Stantcheva, S. 2022. “How To Run Surveys: A Guide to Creating Your Own Identifying Variation and Revealing the Invisible.” National Bureau of Economic Research, Working Paper 30527. https://doi.org/10.3386/w30527 Van Reenen, J. 2015. “Austerity: Growth Costs and Post-Election Plans.”, Center for Economic Performance CEP 2015 Election Analysis Series, #EA020, London School of Economics. Verba, S., K.L. Schlozman and H. Brady. 1995. Voice and Equality: Civic Voluntarism in American Politics (4th ed.). Cambridge M. A.: Harvard University Press. https://doi.org/10.2307/j.ctv1pnc1k7 Walsh, D., R. Dundas, G. McCartney, M. Gibson and R. Seaman. 2022. “Bearing the Burden of Austerity: How Do Changing Mortality Rates in the UK Compare Between Men and Women?”, Journal of Epidemiology Community Health 76 (12): 1027-1033. https://doi.org/10.1136/jech-2022-219645 Wiedemann, A. 2022. “The Electoral Consequences of Household Indebtedness under Austerity.” American Journal of Political Science 00 (0): 1–18. https://doi.org/10.1111/ajps.12708 29 Table 3: Observations excluded from original sample Groups of participants excluded Number Participants with GPS data outside of UK 1 Participants who did not pay attention 2 Participant who did not give the consent to use her/his data for scientific purposes 1 Participants who preferred not to report their ethnicity 14 Total participants excluded 18 Source: authors’ calculations. 30 Table 4: Effect of Exposure to Austerity on political participation Political participation: V oting Protesting Petition against austerity (1) (2) (3) Panel A: Treatment 0.128*** 0.063 0.096*** [0.039] [0.059] [0.024] Controls Yes Yes Yes Control group mean 3.758 2.321 0.395 Obs. 1,476 1,476 1,476 R-squared 0.084 0.184 0.230 Panel B: Treatment 1 (Exposure to Austerity) 0.098** 0.046 0.077*** [0.045] [0.068] [0.028] Treatment 2 (Exposure to Austerity Deprived) 0.157*** 0.081 0.114*** [0.045] [0.068] [0.028] Controls Yes Yes Yes p-value diff t1-t2 0.301 0.914 0.308 Obs. 1,476 1,476 1,476 R-squared 0.088 0.189 0.233 Note: ‘Voting’ refers to the probability that the respondent would ‘Vote for a candidate for public office because of their position on austerity’. Answers range from 1 (definitely would not) to 4 (definitively would). ‘Protesting’ refers to an individual’s willingness to protest against austerity. Answers range from 1 (definitely will not) to 5 (definitively will). ‘Signing a petition AGAINST austerity’ is a dummy variable. Control variables include gender, age bracket, ethnic background, being married, living in the UK, university degree, being in paid employment (full or part-time), income class, trust in government, and being ‘left wing’ or ‘left center’. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.10. Source: authors’ calculations. 31 Table 5: Effect of Exposure to Austerity on preferences for redistribution Preferences for redistribution: Government should reduce income inequality Welfare spending (e.g., income support or child benefits) Social security spending Raising taxes on richer citizens to reduce budget deficit (1) (2) (3) (4) Panel A: Treatment 0.092* 1.771*** 1.072*** 0.102** [0.049] [0.380] [0.306] [0.050] Control group mean 4.144 15.521 12.990 4.208 Obs. 1,476 1,476 1,476 1,476 R-squared 0.192 0.077 0.044 0.171 Panel B: Treatment 1 0.031 1.559*** 1.023*** 0.064 [0.057] [0.439] [0.354] [0.058] Treatment 2 0.154*** 1.980*** 1.121*** 0.139** [0.057] [0.438] [0.353] [0.058] Control group mean 4.144 15.521 12.990 4.208 p-value diff t1-t2 0.031 0.338 0.783 0.193 Obs. 1,476 1,476 1,476 1,476 R-squared 0.195 0.077 0.044 0.172 Note: ‘Government should reduce inequality’ refers to the opinion of participants on the following statement: ‘The government should implement policies to reduce differences in income levels between the rich and the poor?’. Answers range from 1 (strongly disagree) to 5 (strongly agree). Spending variables are continuous, reflecting respondents’ preferred share of total budget being spent on that item. ‘Raising taxes’ refers to the opinion of participants on the following statement: ‘If the UK's national budget deficit is too high, the government should raise taxes on the rich instead of cutting benefits for the poor’. Answers range from 1 (strongly disagree) to 5 (strongly agree). Control variables include gender, age bracket, ethnic background, being married, living in the UK, university degree, being in paid employment (full or part-time), income class, trust in government, being ‘left wing’ or ‘left center’. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.10. Source: authors’ calculations. 32 Table 6: Mediation analysis, overall treatment – summary table of the estimated Equations (3), (4), and (5) V oting Petition AGAINST austerity Effect of treatment on candidate mediator (𝛽) Association of mediator with the outcome variable (𝛿) Effect of treatment on outcome variable (𝛼) % treatment effect on the outcome explained by this mechanism (𝛽* 𝛿󰇜/𝛼 Association of mediator with the outcome variable (𝛿) Effect of treatment on outcome variable (𝛼) % treatment effect on the outcome explained by this mechanism (𝛽* 𝛿󰇜/𝛼 Treatment 0.092* 0.112*** 0.128*** 8.05 0.117*** 0.096*** 11.21 [0.049] [0.034] [0.039] [0.021] [0.024] Treatment 1 0.031 0.112*** 0.106** 9.72 0.117*** 0.082*** 13.13 [0.057] [0.034] [0.045] [0.021] [0.028] Treatment 2 0.154*** 0.112*** 0.153*** 6.73 0.117*** 0.111*** 9.70 [0.057] [0.034] [0.045] [0.021] [0.028] Note: ‘Voting’ refers to the probability that the respondent would ‘Vote for a candidate for public office because of their position on austerity’. Answers range from 1 (definitely would not) to 4 (definitively would). ‘Signing a petition AGAINST austerity’ is a dummy variable. Control variables include gender, age bracket, ethnic background, being married, living in the UK, university degree, being in paid employment (full or part-time), income class, trust in government, and being ‘left wing’ / ‘left center’. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.10. Source: authors’ calculations. 33 Appendix A: Data description This section briefly describes our data and provides some descriptive statistics. Figure A1: Share of citizens who agreed or strongly agreed that the government should act to reduce inequality, 2001 - 2019 Source: authors’ elaboration on the British Election Study (BES) post-electoral survey Table A1: Data sources V ariable Description Source Total financial loss due to austerity Information on total financial loss due to austerity measures was calculated by Beatty and Fothergill (2013;2016) at LAD level as part of a project commissioned by OXFAM and the Joseph Rowntree Foundation Beatty and Fothergill Vote Turnout Data on vote for the Conservative, Labour, and UKIP parties and turnout rates for the period 2000-2019 are obtained from the UK Parliament House of Commons Library. Match between constituency level data and LAD was possible thanks to the work of the Election Centre, affiliated with Nuffield College, Oxford. UK Parliament House of Commons Library Election Centre Protest events Data on protest events for the period 2000-2020 was obtained from GDELT Project, which monitors the world's broadcast, print, and web news from around the world to extrapolate information on different types of events and classify them based on the CAMEO framework. Within this dataset, we identified protest events directed towards the government and that made use of violence. The GDELT Project Appeals Data appeals for political change for the period 2000-2020 was obtained from GDELT Project, which monitors the world's broadcast, print, and web news from around the world to extrapolate information on different types of events and classify them based on the CAMEO framework. The GDELT Project Average income Income data was obtained from the UK Household Longitudinal Survey (UKHLS) for the entire 2000-2020 period. UKHLS is a representative panel survey conducted by UKRI Economic and Research Council, ISER, and University of Essex. Individual-level data was collapsed at LAD-level. UK Household Longitudinal Survey Share of white population We compile information on the population share identifying as ethnically British i.e., either English, Scottish, or Welsh) from the Uk UK Annual Population Survey 34 Annual Population Survey (APS), a representative dataset collected by the Office of National Statistics (ONS). We recode this information into a dummy variable for British White and non-British White ethnic groups and average it at the LDA level to quantify the share of the selfidentified ethnically British White population within each district. Share of population with higher education Data on share of individuals with higher education, considered as individuals with more than secondary education, was obtained at LAD level from UK Annual Population Survey (APS), a representative dataset collected by the Office of National Statistics (ONS) UK Annual Population Survey Share of manufacturing jobs Data on share of individuals working in the manufacturing sector was obtained at LAD level from UK Annual Population Survey (APS), a representative dataset collected by the Office of National Statistics (ONS). UK Annual Population Survey Population Population figures for the year 2000 are used as a weight throughout the analysis. These figures are obtained from Census data estimates obtained from the UK Office of National Statistics UK Census Data Table A2: Descriptive statistics Obs Mean Std. de v Min Max Voting turnout 1,812 65.29 5.83 41.71 81.06 Demand for political reform 7,959 0.09 1.57 0.00 76.00 Protests against the government 7,959 2.40 16.32 0.00 361.00 Austerity 7,770 447.71 120.97 177.00 914.00 (log) average income 7,518 108.14 59.73 6.77 241.11 Share of white population 7,959 8.21 4.62 0.30 17.83 Share of population with higher education 7,203 288.21 189.24 9.80 1692.60 Share of manufacturing jobs 7,182 149.32 105.47 1.80 640.50 35 Appendix B: DD dynamic version As an additional exercise, we estimate a dynamic version of our main specification splitting the period after the implementation of austerity into individual post-period years. The rationale behind is to understand how austerity cuts affected political participation each year after the announcement of these measures. Figure B1 shows that the impact of austerity was positive and statistically significant in the 2015 and 2017 political elections. However, it was no longer statistically significant in the 2019 political elections. Figure B2 shows a positive association between austerity and demands for political reforms only in two years: 2012 and 2018. Both years are related to the initial period of the two austerity waves (2011-2015 and 2016-2019). Figure B3 confirms no association between austerity and protests. Figure B1. Austerity and elections 0.02 .04 .06 Effects on linear prediction 36 Figure B2: Austerity and demand for political reform Figure B3: Austerity and protests -.2 0.2 .4 Effects on linear prediction -.6 -.4 -.2 0.2 Effects on linear prediction 37 Appendix C: Robustness tests Table C1: Excluding Scotland (1) (2) (3) V oter Turnout Appeals for political reform Protests against the government Austerity*post 2010 0.033*** 0.082** 0.008** (0.007) (0.040) (0.004) Observations 1,612 6,384 6,384 R-squared 0.933 0.341 0.233 Controls Yes Yes Yes District FE Yes Yes Yes Region*Year FE Yes Yes Yes Note: Table C1 reports results studying the link between Austerity and political participation over the period 2000 to 2019 excluding Scotland. Austerity*post 2010 refers to the estimated financial losses per working age adult per LAD for the period 2011-2015. The regressions control for the (natural logarithm of) income, the share of white population, the share of the population with higher education and the share of manufacturing jobs per LAD. The regressions also control for local authority district fixed effects and region-year fixed effects. Columns 1-3 are weighted by the LAD adult population in 2000. Standard errors (in parentheses) are clustered at LAD level. *** p<0.01, ** p<0.05, * p<0.1. Table C2: Alternative measures of austerity 2011 – 2015 - outturn 2011 – 2019 - estimated costs (1) (2) (3) (4) (5) (6) V oter Turnout Appeals for political reform Protests against the government V oter Turnout Appeals for political reform Protests against the government Austerity*post 2010 0.021*** 0.099** 0.005 0.026*** 0.110** 0.006 (0.008) (0.043) (0.004) (0.008) (0.047) (0.004) Observations 1,754 6,930 6,930 1,754 6,930 6,930 R-squared 0.932 0.363 0.239 0.933 0.363 0.239 Controls Yes Yes Yes Yes Yes Yes District FE Yes Yes Yes Yes Yes Yes Region*Year FE Yes Yes Yes Yes Yes Yes Note: Table C2 reports results studying the link between Austerity and political participation over the period 2000 to 2019 using alternative measures of Austerity. The regressions control for the (natural logarithm of) income, the share of white population, the share of the population with higher education and the share of manufacturing jobs per LDA. The regressions also control for local authority district fixed effects and region-year fixed effects. Columns 1-6 are weighted by the LAD adult population in 2000. Standard errors (in parentheses) are clustered at LAD level. *** p<0.01, ** p<0.05, * p<0.1. 38 Table C3: Alternative dependent variables (1) (2) (3) V oting Appeals for political reform Protests Austerity*post 2010 0.098** 0.036*** 0.018 (0.042) (0.014) (0.026) Observations 4,457 1,974 6,909 R-squared 0.384 0.077 0.497 Controls Yes Yes Yes District FE Yes Yes Yes Region*Year FE Yes Yes Yes Note: Table C3 reports results studying the link between Austerity and political participation over the period 2000 to 2019. Voting refers to information on whether individuals voted in past general election. Appeals for political reform refers to information on whether individuals signed a petition (not on the internet). Protests is a dummy variable with value one if a protest occurred in the LAD. Austerity*post 2010 refers to the estimated financial losses per working age adult per LAD for the period 2011-2015. The regressions control for the (natural logarithm of) income, the share of white population, the share of the population with higher education and the share of manufacturing jobs per LDA. The regressions also control for local authority district fixed effects and region-year fixed effects. Columns 1-3 are weighted by the LAD adult population in 2000. Standard errors (in parentheses) are clustered at LAD level. *** p<0.01, ** p<0.05, * p<0.1. Our alternative measures for voting and appealing are from the UK Household Longitudinal Survey (UKHLS) and the British Election Study (BES). These two surveys collect individual-level data about the social, economic, and political views and circumstances of British society. Because of social desirability bias, these surveys may inaccurately represent the population’s political views, in particular if participants hold more controversial political views (Funk 2016). Our alternative measure for protests is obtained from the Mass Mobilization Protest (MMP) database from Binghamton University (Clark and Regan 2018). This data explicitly focuses on protests against governments and is collected and revised by researchers, as opposed to the algorithm-based approach of GDELT. MPP, however, records considerably fewer protests and might underestimate protests activity. Furthermore, it provides inconsistent information on protest locations at the subnational level. To obtain precise location information, MPP relies on an API called Open Cage (https://opencagedata.com/). The API uses available information to calculate GPS coordinates that are as precise as possible for the protest, providing information on the quality of the coordinate. We only select observations with the highest quality possible and end up with a total of 578 protest events, of which 393 happened between the year 2000 and 2020. 45 Please watch this video on weather forecast technology carefully. Make sure your computer audio is working, so you can follow what is being said in the video. You will be able to advance the survey once the video is finished. 24.1 Choose all that apply. Data from space satellites... ...help us see the weather before it reaches us; ...cannot be used for forecasting weather; ...are combined with other data to make weather predictions; ...come from stationary and earth-orbiting satellites; ...come from stationary satellites; ...come from earth-orbiting satellites. Please watch this video on UK austerity policies and their economic and social impact carefully. Make sure your computer audio is working, so you can follow what is being said in the video. You will be able to advance the survey once the video is finished. 24.2 Choose all that apply. Austerity implied that the UK government: cut welfare spending; raised taxes; lowered taxes; increased welfare spending 24.3 Compared to other countries UK austerity policies were: Strict; less strict; about the same Please watch this video on UK austerity policies and their economic and social impact. Make sure your computer audio is working, so you can follow what is being said in the video. You will be able to advance the survey once the video is finished. 24.4 Choose all that apply. Austerity implied that the UK government: cut welfare spending; raised taxes; lowered taxes; increased welfare spending 24.5 Compared to other countries UK austerity policies were: Strict; less strict; about the same --- 25. The government should implement policies to reduce differences in income levels between the rich and the poor. Strongly agree; agree; Neither agree nor disagree; disagree; strongly disagree 26. Suppose your can decide over total UK government spending next year. How do you want to divide the budget in percentages between the following categories? (The total must sum up to 100%) Health: _______ ; Education: _______ ; Defense and national security: _______ ; Welfare spending (e.g., income support or child benefits): _______ ; Social security: _______ ; Public infrastructure: _______ ; Total: ________ 46 27. Please indicate whether you agree or disagree with the following statement: ‘If the UK's national budget deficit is too high, the government should raise taxes on the rich instead of cutting benefits for the poor.’ Strongly agree; agree; Neither agree nor disagree; disagree; strongly disagree 28. Now that you have learned about austerity. How likely would you be to do each of the following things? Definitively would; Probably would; Probably would not; Definitely would not - Vote for a candidate for public office because of their position on austerity - Write letters, email, or phone public officials about austerity policies - Meet with an elected official or their staff about austerity policies - Attend a public speech against austerity policies - Donate money to a non profit organization helping disadvantaged households in the UK - Volunteer time to a non profit organization helping disadvantaged households in the UK 29. If a peaceful protest against austerity policies was being held this week close to where you live, would you participate? Definitely will not; Probably will not; Might or might not; Probably will; Definitely will 30. You can also take action by signing a petition. We will submit each petition to two Members of Parliament who have either spoken for or against austerity policies in parliament. We will not tell them your name, just how many people in our study support either of the two petitions below. We will send you proof of the petition's submission to the corresponding Member of Parliament in the next few weeks. Would you like to sign a petition? - I want to sign the following petition AGAINST austerity policies. ‘Austerity policies affect all segments of the population and particularly hurt the most vulnerable groups. Spending cuts should be stopped. I am against austerity policies.’ - I want to sign the following petition FOR austerity policies. ‘Cuts are necessary to ensure economic sustainability. Benefits should be reserved for people who really need them such as old people and children. I support austerity policies.’ I do not want to sign either of these petitions. 47 Appendix F: Informational treatments VIDEO 1: Treatment austerity only What is the U.K. austerity program? The global financial crisis of 2007 and of 2009 led to a severe economic crisis in the U.K. and many other countries. As a response, the U.K. government launched an austerity program in 2010. This U.K. austerity program was adopted to deal with the consequences of the economic crisis. Specifically, it was implemented to reduce budget deficits and public debt by cutting government spending and welfare benefits and raising taxes to increase revenue. So what was the impact of the U.K. austerity program on the economy and society? Compared to other countries, U.K. austerity policies were very strict. For example, between 2010 and 2018, the U.K. government reduced spending in welfare payments, housing subsidies and social services across society by more than £30 billion. This translates to 15% less day to day government spending per citizen compared to the time before austerity. Still, the U.K. Treasury pronounced austerity a success as the economy has grown by 18.3% since 2010, faster than France, Italy and Japan, whose austerity policies were not a strict. So why were austerity policies widely protested by the public? Researchers and experts, including British economists, point to the damage that austerity has dealt to society. The use of foodbanks almost doubled between 2013 and 2017. Child poverty, including in families with working parents, rose to its highest level since before the Second World War. Infant mortality increased for the first time in two generations. What is more, lingering deep cuts to youth council services since 2010 are being linked to increasing crime and riots across different parts of the country. Under austerity, life expectancy in the UK fell and this trend has only recently reversed. Despite the formal announcement of the end of austerity in October 2018, its effects continue to reverberate throughout the country. All of this indicates that austerity has been compromising living standards and progress that the UK society has made over the past decades. VIDEO 2: Treatment austerity and deprivation What is the U.K. austerity program? The global financial crisis of 2007 and of 2009 led to a severe economic crisis in the U.K. and many other countries. As a response, the U.K. government launched an austerity programme in 2010. This U.K. austerity program was adopted to deal with the consequences of the economic crisis. Specifically, it was implemented to reduce budget deficits and public debt by cutting government spending and welfare benefits and raising taxes to increase revenue. So what was the impact of the U.K. austerity program on the economy and society? Compared to other countries, U.K. austerity policies were very strict. For example, between 2010 and 2018, the U.K. government reduced spending and welfare payments, housing subsidies and social services across society by more than 30 billion pounds. This translates to 15% less day to day government spending per citizen compared to the time before austerity. Still, the U.K. Treasury pronounced austerity a success as the economy has grown by 18.3% since 2010, faster than France, Italy and Japan, whose austerity policies were not as strict. So why were austerity policies widely protested by the public? Researchers and experts, including British 48 economists, point to the damage that austerity has dealt to society. The use of foodbanks almost doubled between 2013 and 2017. Child poverty, including in families with working parents, rose to its highest level since before the Second World War. Infant mortality increased for the first time in two generations. What is more, lingering deep cuts to youth council services since 2010 are being linked to increasing crime and riots across different parts of the country. Under austerity, life expectancy in the UK fell and this trend has only recently reversed, despite the formal announcement of the end of austerity in October 2018. Its effects continue to reverberate throughout the country. All of this indicates that austerity has been compromising living standards and progress that the UK society has made over the past decades. Even worse, poorer UK citizens living in the most deprived local authorities are also the ones hit hardest by the billions in austerity cuts. There are countless examples. Rhea, a working single mother of three, was driven to the brink of homelessness due to reduced housing benefits. Changing eligibility requirements and stretched local authorities led to her case being unresolved for weeks. George, a single father and bus driver, was hit hard by reduced welfare support and delays in it being paid. This was enough to throw him into a vicious cycle of debt. Lorna, a mother and school dinner lady, could not send her son into school for a couple of days as she couldn't afford a packed lunch for him without government support available. Lorna decided to visit the foodbank for the first time. This made her feel very guilty and ashamed. These are just a few of many stories of poor UK citizens hit hard by austerity. 49 Appendix G: Quality tests Table G1 reports the result of the balancing test. It shows that the comparability of experimental groups is not affected by the exclusions of 18 observations (as explained in section Data quality and descriptive statistics) Table G1. Balancing test (1) (2) (3) Treatment Treatment 1 Treatment 2 Female -0.0109 -0.0185 0.00765 (0.0248) (0.0248) (0.0248) Age group 0.0166* 0.0121 0.00451 (0.00906) (0.00904) (0.00906) Married -0.0211 -0.0437 0.0225 (0.0273) (0.0272) (0.0273) Asian (ethnicity) -0.00179 0.00322 -0.00501 (0.0485) (0.0484) (0.0485) Mixed Background (ethnicity) 0.00326 -0.0484 0.0517 (0.0525) (0.0524) (0.0525) UK resident 0.0115 -0.0391 0.0506 (0.0500) (0.0499) (0.0500) University 0.00716 -0.0390 0.0462* (0.0260) (0.0260) (0.0261) Income group 0.00486 0.00501 -0.000149 (0.00558) (0.00557) (0.00558) Fulltime employment -0.0443* 0.00259 -0.0469* (0.0265) (0.0264) (0.0265) Trust in government -0.00618 -0.00879 0.00261 (0.0158) (0.0158) (0.0158) Left leaning 0.0189 0.0173 0.00155 (0.0261) (0.0260) (0.0261) Constant 0.623*** 0.378*** 0.245*** (0.0817) (0.0816) (0.0818) Observations 1,476 1,476 1,476 R-squared 0.006 0.005 0.007 We run an additional test to check the quality of our data. Immediately after the treatment videos, we asked participants two multiple choice questions to test their knowledge about the video content. First, we asked about the type of measures which were implemented by the UK government. Second, we asked about the stringency of UK austerity policies compared to other European countries. About 99% of respondents gave at least one correct response to the first question; 96% correctly answered the second question. Our results hold when we run a robustness 50 check, we run our baselines regressions on the sample of participants who gave at least one correct response to the first question and correctly answered the second (Table G2 in Appendix G). Table G2. Main results excluding participants who took longer to complete the survey (Panel A) and those who did not reply correctly to at least one question (Panel B) V oting Protesting Petition AGAINST austerity (1) (2) (3) Panel A Treatment 0.122*** 0.062 0.092*** [0.040] [0.059] [0.025] Control mean 3.758 2.321 0.395 Obs. 1,449 1,449 1,449 R-squared 0.087 0.184 0.232 Panel B Treatment 0.126*** 0.059 0.103*** [0.040] [0.059] [0.025] Control mean 3.758 2.321 0.395 Obs. 1,436 1,436 1,436 R-squared 0.087 0.189 0.228 Note: ‘Voting’ refers to the probability that the respondent would be to ‘Vote for a candidate for public office because of their position on austerity’. Answers could range from 1 (Definitely would not) to 4 (Definitively would). ‘Protesting’ refers to the willing of the respondent to participate to a protest against austerity policies. Answers could range from 1 (Definitely will not) to 5 (Definitively will). ‘Signing a petition AGAINST austerity’ and ‘Signing a petition FOR austerity’ is a dummy variable. Standard errors in parentheses. *** p<0.01, ** p<0.05. 51 Appendix H: Prolific sample Table H1. Descriptive statistics V ariable Share Female 0.510 Age 18-24 0.096 25-34 0.198 35-44 0.190 45-54 0.161 55-64 0.234 65-74 0.111 75-84 0.011 UK citizen 0.929 Ethnic origin White 0.863 Asian 0.075 Mixed background 0.062 Income Less than 15,000 0.131 15,000 - 24,999 0.182 25,000 - 34,999 0.207 35,000 - 44,999 0.150 45,000 - 54,999 0.112 55,000 - 64,999 0.084 65,000 - 74,999 0.049 75,000 - 84,999 0.028 85,000 - 94,999 0.018 95,000 - 104,999 0.018 105,000 - 144,999 0.015 More than 145,000 0.007 In paid employment 0.557 Married 0.447 University degree or higher 0.576 Left leaning 0.474 Source: authors’ elaboration on Prolific survey 52 Appendix I: Additional results online experiment Table I1: Exposure to austerity on Political participation V oting Protesting Petition AGAINST austerity (1) (2) (3) Panel A: without controls Treatment 0.123*** 0.056 0.099*** [0.040] [0.064] [0.027] Control mean 3.758 2.321 0.395 Obs. 1,490 1,490 1,490 R-squared 0.006 0.001 0.009 Panel B: without trust and political controls Treatment 0.134*** 0.077 0.102*** [0.040] [0.063] [0.027] Control mean 3.758 2.321 0.395 Obs. 1,476 1,476 1,476 R-squared 0.023 0.047 0.032 Note: ‘Voting’ refers to the probability that the respondent would be to ‘Vote for a candidate for public office because of their position on austerity’. Answers could range from 1 (Definitely would not) to 4 (Definitively would). ‘Protesting’ refers to the willing of the respondent to participate to a protest against austerity policies. Answers could range from 1 (Definitely will not) to 5 (Definitively will). ‘Signing a petition AGAINST austerity’ is adummy variable. Standard errors in parentheses. *** p<0.01, ** p<0.05. 53 Table I2: Exposure to austerity on the Mediators Government should reduce inequality Welfare spending Social security spending Raising taxes (1) (2) (3) (4) Panel A: without controls Treatment 0.088 1.784*** 1.154*** 0.099* [0.054] [0.390] [0.308] [0.054] Control mean 4.144 15.521 12.990 4.208 Obs. 1,490 1,490 1,490 1,490 R-squared 0.002 0.014 0.009 0.002 Panel B: without trust and political controls Treatment 0.104* 1.799*** 1.101*** 0.112** [0.054] [0.387] [0.309] [0.054] Control mean 4.144 15.521 12.990 4.208 Obs. 1,476 1,476 1,476 1,476 R-squared 0.033 0.042 0.021 0.027 Note: ‘Government should reduce inequality’ is a discrete variable ranging from 1 (strongly disagree) to 5 (strongly agree). ‘Raising Taxes’ refers to the opinion of participants on the following statement: ‘if the UK's national budget deficit is too high, the government should raise taxes on the rich instead of cutting benefits for the poor’. Answers could range from 1 (strongly disagree) to 5 (strongly agree). Spending variables are continuous, reflecting respondents’ preferred share of total budget being spent on that particular item. Standard errors in parentheses. *** p<0.01, ** p<0.05. 54 Appendix J: Mediation analysis Table J1: Exposure to Austerity on redistributive preferences, pooled treatments Preferences for redistribution: Government should reduce income inequality Welfare spending Social security spending Raising taxes on richer citizens to reduce budget deficit (1) (2) (3) (4) Panel A: Treatment 0.092* 1.771*** 1.072*** 0.102** [0.049] [0.380] [0.306] [0.050] Control group mean 4.144 15.521 12.990 4.208 Obs. 1,476 1,476 1,476 1,476 R-squared 0.192 0.077 0.044 0.171 Panel B: Treatment 1 0.031 1.559*** 1.023*** 0.064 [0.057] [0.439] [0.354] [0.058] Treatment 2 0.154*** 1.980*** 1.121*** 0.139** [0.057] [0.438] [0.353] [0.058] Control group mean 4.144 15.521 12.990 4.208 p-value diff t1-t2 0.031 0.338 0.783 0.193 Obs. 1,476 1,476 1,476 1,476 R-squared 0.195 0.077 0.044 0.172 Note: ‘Government should reduce inequality’ is a discrete variable ranging from 1 (strongly disagree) to 5 (strongly agree). ‘Raising Taxes’ refers to the opinion of participants on the following statement: ‘if the UK's national budget deficit is too high, the government should raise taxes on the rich instead of cutting benefits for the poor’. Answers range from 1 (strongly disagree) to 5 (strongly agree). Spending variables are continuous, reflecting respondents’ preferred share of total budget being spent on that particular item. Control variables include gender, age bracket, ethnic background, being married, living in the UK, university degree, being in paid employment (full or part-time), income class, trust in government, being ‘left wing’ or ‘left center’. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.10. The UNU-MERIT WORKING Paper Series 2023-01 Can international mobility shape students' attitudes toward inequality? The Brazilian case by Cintia Denise Granja, Fabiana Visentin and Ana Maria Carneiro 2023-02 Demand-led industrialisation policy in a dual-sector small open economy by Önder Nomaler, Danilo Spinola and Bart Verspagen 2023-03 Reshoring, nearshoring and developing countries: Readiness and implications for Latin America by Carlo Pietrobelli and Cecilia Seri 2023-04 The role of product digitization for productivity: Evidence from web-scraping European high-tech company websites by Torben Schubert, Sajad Ashouri, Matthias Deschryvere, Angela Jäger, Fabiana Visentin, Scott Cunningham, Arash Hajikhani, Lukas Pukelis and Arho Suominen 2023-05 More than a feeling: A global economic valuation of subjective wellbeing damages resulting from rising temperatures by Stephan Dietrich and Stafford Nichols 2023-06 Was Robert Gibrat right? A test based on the graphical model methodology by Marco Guerzoni, Luigi Riso and Marco Vivarelli 2023-07 Predicting social assistance beneficiaries: On the social welfare damage of data biases by Stephan Dietrich, Daniele Malerba and Franziska Gassmann 2023-08 Related or unrelated diversification: What is smart specialization? By Önder Nomaler and Bart Verspagen 2023-09 Breach of academic values and digital deviant behaviour: The case of Sci-Hub by Giulia Rossello and Arianna Martinelli 2023-10 The effect of lobbies’ narratives on academics' perceptions of scientific publishing: An information provision experiment by Giulia Rossello and Arianna Martinelli 2023-11 Making impact with agricultural development projects: The use of innovative machine learning methodology to understand the development aid field by Lindsey Moore, Mindel van de Laar, Pui Hang Wong and Cathal O’Donoghue 2023-12 Green windows of opportunity in the Global South by Rasmus Lema and Roberta Rabellotti 2023-13 The green and digital transition in manufacturing global value chains in latecomer countries by Rasmus Lema and Roberta Rabellotti 2023-14 Impact of prestigious-STEM Education of corporate board members on innovation effort: Evidence from India by Rituparna Kaushik, Sourabh Bikas Paul and Danilo Spinola 2023-15 Reducing environmental impact through shared ownership: A model of consumer behaviour by Francesco Pasimeni and Tommaso Ciarli 2023-16 Welfare losses, preferences for redistribution, and political participation: Evidence from the United Kingdom’s age of austerity by Patricia Justino, Bruno Martorano and Laura Metzger