Research

Publications





Working Papers 

Subsidies and Democratic Backsliding in the EU: Evidence from Hungary

with Tom Pepinsky and Ádám Reiff

This paper studies the attitudes and policy preference of the mass public following a large-scale, unconditional government spending program. In 2022, Hungary’s conservative government implemented a large-scale pre-election spending program, paying out the country’s largest pension bonus and giving families a huge tax refund. Beyond the size of the payment, these measures provided unconditional cash transfers to a large share of voters and rewarded the relatively better off people. How do these benefits affect citizens’ political participation and party preference? How do the mass public view the appropriateness of supporting a party based solely on material subsidies? We analyze the effect of the family tax refund and the 13th month pension payment on recipients’ party preference. At the same time, while receiving these handouts was not illegal, people are still reluctant to admit that they support a party based on material rewards. The resulting reluctance of admitting allegedly inappropriate behavior is called social desirability bias. To elicit unbiased answers about this sensitive behavior and to understand how much of a problem is sensitivity bias in this case, we employ a list experiment technique. Our results suggest that these subsidies worked mainly by demobilizing voters who might have opposed the incumbent party. In particular, family tax refund recipients with primary education or living in rural areas, and the pension recipients living in the capital city were less likely to vote for the opposition coalition than non-recipient voters with similar socio-demographic characteristics. The findings also reveal that the material rewards influenced the party preference of around 20% of the incumbent voters.


New Parties and Their Media Coverage: An Experimental Study of Media Practitioners in Italy

with Ádám Reiff, Joost van Spanje and Alessio Scopelliti

Studies on news values have provided many insights into which party messages get reported in the media. Nonetheless, less attention has been paid on whether and when new party’s messages are reported. In this study, we use the concept of news values to examine how journalists from four Italian regions perceive the newsworthiness of party messages. Using conjoint survey experiment, fictional party press releases are used to test for the influence of five important news values and how the importance of these news values differ across new and established parties. Very preliminary findings suggest that the newsworthiness of a press release issued by a new party significantly differs from those issued by an established party.

Targeted Government Subsidies and Electoral Support – The Hungarian Village Programme and Political Support in Hungary 

with Ádám Reiff

This paper examines the effect of two subsidy programmes targeting rural areas on electoral support for Hungary’s right-wing ruling party (Fidesz). The first programme, the Rural Family Housing Allowance Program (programmatic distributive strategy), was given directly to individuals living in eligible small settlements. The second programme, the Hungarian Village Program (pork-barrel politics), provided an indirect support to individuals by proffering public resources to the local governments in eligible small settlements. The programmes were introduced in July 2019, shortly after the EP elections in May 2019 and just before the local elections in October 2019. We rely on a difference in differences estimation strategy by comparing average changes in Fidesz vote share from before the introduction of the policies to after it between eligible and non-eligible settlements. We find that Fidesz vote share increases by 4.3 percentage points in settlements that were eligible for both programs relative to non-eligible settlements. Additionally, we measure the effect of the amount of subsidies on Fidesz vote share and find that an additional HUF 10,000 (approximately 30 USD) HVP subsidy increases Fidesz vote share by around 1.2 percentage points (relative to Fidesz vote share in those settlements that did not receive this additional subsidy). In contrast, the amount of Rural CSOK subsidy does not have a significant effect on Fidesz vote share. We suggest that Fidesz targeted its core supporters with these policies and mobilised core voters (rather than pursued swing voters).


The Short-term Health Effects of Conflict: Evidence from Sub-Saharan Africa 

with Caitlin Brown

This paper investigates how exposure to different types of conflict affects health outcome for individual in general and for women and children in particular. Over the period 2003 – 2018, health outcomes of individuals from the Demographic and Health Surveys (DHS) database are matched with conflict events from the Armed Conflict Location and Event Database (ACLED) within a 241 km (150 miles) radius that occurred 120 days before the date of DHS interview. The study examine the short-term effect of conflict on various health outcomes (on height for age z-scores, weight-for-age z-scores and weight-for-age z-scores) in 33 sub-Saharan African countries. The main goal of the study is to understand (1) how conflict affects health outcome for women and children in Sub-Saharan Africa; (2) whether household members are affected differently within the household; (3) through what mechanisms these effects are operating. 


How Do Citizens Respond Politically to the Local Economy? Evidence from Hungarian Administrative and Survey Data 

with Ádám Reiff

How do citizens form perceptions of a complex and abstract economy and how is perception of the economy translated into vote choice? Recent studies of economic voting have focused on the role of the local economy, but with inconclusive results. A large literature finds that the local economic milieu has a substantial effect on incumbent support – incumbent politicians are rewarded for strong local economic performance and punished for weak performance. There are, however, reasons to expect that individuals’ economic perceptions are endogenous to vote choice, meaning that existing cross-sectional models cannot provide a valid test of the causal local economic voting claim. In this study, we use a novel dataset and link a nationally-representative election survey with more than 75,500 respondents to extraordinary granular local economic administrative data measured without sampling error. Using an instrumental variable approach, we assess the effect of sociotropic evaluations on the decision to vote for the incumbent government in Hungary by exploiting variations in objective local economic measures. Findings show that changes in unemployment rate as well as changes in income level have strong predictive power on individuals’ sociotropic evaluations with changes in income having slightly stronger effect; while the effect of sociotropic evaluations on party preferences are stronger during the peak of the Great Recession. The study thus provides a framework for understanding when citizens respond politically to the local economy.