Research project grants
Climate Risk, Land Loss, and Migration: Evidence from a Quasi-Experiment in Bangladesh (Swiss National Science Foundation, 01.09.2019 – 31.08.2023, 442’284 CHF) (Co-led with V. Koubi) [Abstract]
Global climate change is one of the most important and severe challenges the international community has ever faced. Existing evidence shows that it will have far-reaching repercussions for ecosystems and humans alike. Moreover, climate change is expected to induce mass-population dislocations, i.e., migration, due to droughts, sea level rise, or extreme weather events, such as stronger and more frequent storms or floods, particularly in developing countries with low capacity to protect themselves and adapt to climate change. However, recent studies on climate change induced human displacement do not account for the possibility that people might adapt to changing climatic conditions. This is particularly relevant to slow-onset environmental changes, such as sea-level rise, where individuals and societies can anticipate such changes and take precautionary measures. With this project, we aim at contributing to a better understanding of whether, when, and how environmental changes lead to human migration. First, we will offer a theoretical micro-foundation for the environment-migration nexus that highlights the different (i.e., behavioral, structural, and environmental) factors that induce people to migrate or stay. In particular, our framework proposes that there are multiple drivers behind decisions to migrate and that environmental changes are just one of them, which can have both direct effects on migration as well as indirect ones through impacts on other factors, such as individual/household economic well-being. Consequently, our framework seeks to bring analytical rigor to a field that has been dominated by unsubstantiated and casual empiricism. Second, we seek to provide credible empirical inferences concerning rates of migration due to livelihood losses caused by environmental changes. Since human exposure to environmental stress, whether induced by climatic changes or other factors, is non-randomly assigned, using observational data to empirically examine environmental migration is challenging. To cope with this challenge, we will focus on a particular case and develop risk maps for riverbank erosion along the Jamuna River in Bangladesh, and we will collect micro-level data to compare affected and unaffected populations at a similar baseline risk, as well as migrants and non-migrants from the very same area. The aim is to isolate the causal effect of deteriorating environmental livelihoods on migration.The project will not only identify the effect of environmental stress on migration behavior of individuals and communities in a particular case. It will also provide valuable insights of broader relevance into whether and how societies react, or could react, to slow-onset climatic changes such as se-level rise, drought, and soil/water salinity. Moreover, the methodology developed in the project can be applied in other cases and can inform prediction models of future climate-induced migration rates. The findings can be utilized by institutional actors at both local and international levels when seeking to identify policy options to increase the adaptive capacity of populations vulnerable to climatic changes.
The Comparative Legitimacy of Arms Exports in Top-Exporting EU and NATO Countries (German Foundation for Peace Research, 01.12.2022 – 31.05.2024, 150’000 € (Co-PI, with P. Thurner)
Einstellungen zu Waffenhandel in Deutschland und Frankreich – A Conjoint Experiment on the Comparative Legitimacy of Arms Exports in Germany and France (German Foundation for Peace Research, 01.06.2020 – 31.05.2021, 24’000 € (Co-PI, with P. Thurner)
The export of arms is severely contested between the German political parties and in the German public. A major argument by opposing civil society groups and mainly leftist parties is the reference to the supposed implications of such transfers: the triggering of civil and international conflict, the prolongation of such conflicts and their aggravation in terms of human losses, the violations of human rights, and the stabilization of non-democratic regimes. It is open, however, whether the political discussion of such arguments is also reflected in citizens’ attitudes towards arms exports. There is also no research on the question whether such attitudes differ systematically across countries. E.g., it seems that other Western democracies with a dominant role in the international arms transfer system like the US, the UK and France are less strongly opposing such exports. We want to provide a thorough empirical foundation for these anecdotal claims which is currently lacking in both scientific and public debate. Thus, a cross-country comparison of voter reactions towards arms exports is especially important for testing the assumption of the presence of a German specificity, but also for the internal reliability of alliances like NATO, or of a recently proposed European Defense Policy including the respective research and development initiatives like the Permanent Structured Cooperation (PESCO). If decisions to transfers arms to countries like Saudi-Arabia, to the Kurds etc. cannot be supported by the German government due specific attitudes in its civil society, this will have consequences for the design of future cooperative defense and security regimes. In the case of France and Germany, open disputes over principles of the exports of jointly developed weapons have even led both countries to settle their conflict in the their Aachen Treaty in 2019, and in more detail in a Franco-German agreement on export controls in October 2019 (see Décret n° 2019-1168 du 13 novembre 2019). This pilot project seeks to provide answers to these essential questions building on an innovative methodical approach. It will use conjoint experiments to be implemented in two selected countries – Germany and France. According to SIPRIs 2019 Weapons transfer statistics, these two countries belong to the top 5 group of exporters of major weapons. In France, these exports are rather seldom a topic of political debates, in Germany that’s repeatedly the contrary.
Conjoint experiments allow to implement an experimental setting within a survey format. Respondents are several times confronted with multidimensional hypothetical decisions between choice sets which differ according to several dimensions (choice attributes). These decision tasks are designed in a way that they mimick concrete policy design options. Respondents are then asked to rate the individual choices and they select their preferred option. Attribute levels, i.e. the concrete description of the situation, are randomly assigned within choice sets. This enables the researcher to identify how choice characteristics causally affect both the rating and choice probability of a policy package in a within subject and between subject-design. We can thereby determine the causal effect of the manipulated dimensions of these scenarios. The project will for the first time focus on the comparative relevance of moral, economic and security aspects on the assessment of the legitimacy of weapons exports in Germany and France. It will derive value trade-offs between the perceived economic welfare impact (jobs, innovation etc.) and normative considerations (risk or presence of conflicts, human rights violations, regime characteristics of the importer). We expect these latter aspects to decrease the acceptance of arms exports in general. However, we anticipate these effects to be smaller or inexistent (ceteris paribus) in the case of France due to a different political history and culture. If this expectation is refuted, than it is rather institutional background conditions like the electoral system and the consequence of a low politicization in such settings of party competition. The pilot character of this study would may invite us to extend this design to more countries (especially to France and to the US) in later periods, and to elaborate in a more detailed way on the time- and context dependent interplay between economic and normative considerations in weapons exports.
Working papers (submitted/conference drafts)
Footprint shifting, concern for trade impacts and trade policy preferences (with D. Presberger, F. Quoß, and T. Bernauer) (R&R)
[Link to Manuscript] [Abstract]
Vastly increased international trade over the past few decades has resulted in an ever larger geographical spread in the environmental impacts of local consumption. Particularly in the case of high-income countries, a large share of their total environmental footprint of local consumption now materializes in places far beyond the respective national border. On the presumption that democratic policy-makers should, and often do, act in line with prevailing public opinion we examine whether currently weak policies addressing consumption-based environmental impacts abroad may reflect a knowledge gap amongst citizens, and how closing this knowledge gap would affect policy preferences concerning the greening of international supply chains. We do so based on an experiment, embedded in a large representative survey (N=8’000) in Switzerland, a high-income country with a very large extraterritorial environmental footprint. The main finding is that there is a major knowledge gap amongst the mass public in this area, and that this gap can be closed. However, closing the knowledge gap does not lead to a significant change in policy preferences in favor of reducing the global environmental footprint of local consumption. This points to major policy challenges in trying to mitigate problems of environmental impact shifting in the global economy.
Deontological and Consequentialist Preferences Towards Arms Exports — A Comparative Conjoint Experiment in France and Germany (with M. Freitag, and P. Thurner) (under review)
[Link to Manuscript] [Pre-Registration] [Abstract]
Despite fierce politicization and heated public debates in arms-exporting democracies, systematic research on mass public preferences on arms trade is lacking. Combining political economy models of arms trade with the literatures on trade preferences and foreign policy attitudes, we argue that citizens trade off economic incentives, strategic interests and moral considerations when assessing arms trade and that deeply rooted `strategic cultures’ lead to differences in citizen preferences between countries. To derive the implicit weighting of different features of arms trade, we draw on population-representative conjoint survey experiments (N=6,617), fielded in November/December 2020 in two of the global top-5 exporting countries of major arms: Germany and France. We find that both country populations show structured preferences towards arms exports which predominantly center around their moral repercussions. However, German respondents place more weight on moral consequences and, compared to French respondents, a larger share is in fundamental opposition. We conclude that these diverging preferences potentially conflict with plans of a common European defense and security policy.
Operationalisation matters: Weather extremes as noisy natural experiment show no influence on political attitudes (with F. Quoß) (under review) [Link to Manuscript] [Abstract]
Accelerating climate change increases the importance of mass public support for mitigation policies. While the personal experience of weather extremes could theoretically affect citizens’ attitudes and preferences, prior empirical evidence is inconclusive. We propose that this is the case for three reasons: firstly, the causal link between the direct personal experience of extreme weather and changes in attitudes is structured along a chain of four intermediate steps which are not always clearly separated or tested. Secondly, researchers need to measure „weather“ to study its consequences, but there is an extensive variety of potential measures available. This leads to vast researcher degrees of freedom. Thirdly, there is a large variety in research designs applied. Testing 34 different weather operationalisations in a specification curve analysis with Swiss geo-coded panel data, we find that singular specifications could support arbitrary conclusions in any direction, while the whole set indicates robust null effects.
Do distributional consequences affect public goods provision? Insights from 5G antenna placement in Switzerland (with F. Quoß, and T. Bernauer) (under review)
[Link to Manuscript] [Pre-Registration] [Abstract]
Distributional implications of public goods provision may affect the ability of societies to provide these. Particularly, localized provision costs may result in opposition in the vicinity of provision sites, reducing provision levels and/or efficiency (“not-in-my-backyard” (NIMBY) challenge). We examine mass public support on policy provision in this regard, focusing on 5G, the latest technology standard for mobile data transmission, and the placement of 5G antennas in particular. Based on survey experiments with a geo-coded representative sample of over 5’000 residents of Switzerland, revealing real-world antenna locations to respondents, we find that NIMBYism plays a role for individual attitudes/policy preference formation towards 5G expansion. NIMBYism also affects the (stated) propensity to engage in political action against 5G antennas, irrespective of monetary costs. Finally, NIMBYism can be mitigated when citizens decide under a veil of ignorance on a feasible distribution of siting locations, leaving actual site choice to be a technical process.
Ordering effects vs. cognitive burden – How should we structure attributes in conjoint experiments (with M. Freitag, and P. Thurner) (under review)
[Pre-Registration] [Link to Manuscript] [Abstract]
Conjoint experiments have become very popular in recent years in political science and related disciplines. They offer a flexible way to elicit population preferences on complex choice tasks. We investigate whether we can improve citizens‘ survey experience, and ultimately choice quality, by diverging from the current standard of randomized conjoint attribute ordering. Such random ordering guarantees that any potential bias from attribute order cancels out on average. However, in situations with many attributes (> 4) this may unnecessarily increase cognitive demand for respondents, as attributes which belong together in terms of content are presented scattered across the choice table. Hence, we investigate experimentally whether purposeful ordering (where the researcher chooses a display order with „theoretically important“ dimensions first, i.e. dimensions (s)he wishes the respondent to process first) or block-randomized ordering (where the researcher displays attributes belonging to the same theoretical concept in randomized bundles) affects survey experience, response time and choice itself, as compared to a fully randomized ordering of attributes. Drawing on a complex conjoint choice design with nine attributes and 6,600 respondents (and hence sufficient power) from Germany and France we find that neither block-randomization, nor full randomization nor purposeful ordering affects self-reported survey experience, choice task timing or attribute weighting, not even systematically for high vs. low educated respondents where cognitive burden effects could most likely be expected. To our knowledge, we thereby provide the first systematic empirical evidence that ordering effects are likely of low relevance in conjoint choice experiments, and that the trade-off between cognitive burden and ordering effects is minimal from the perspective of respondents, at least for our substance matter.
Selecting Good Types or Holding Incumbents Accountable? Evidence from Reoccurring Floods
A growing literature draws on natural disasters to assess how voters hold governments accountable and finds that good management benefits incumbents. This suggests that voters see disaster management as a test case that acts as an information shock, revealing unobserved incumbent quality. Theoretically, this depends on the perception and attribution of incumbent responsibility. Additionally, past experience with exposure should moderate voters’ appraisal, as should the current level of esteem the incumbent holds. I provide evidence that the electoral response to disaster management is heterogeneous along these dimensions. Drawing on the exceptional case of four centennial floods in Germany, occurring within a decade and right before elections, I show that exposure leads to vote gains for federal and state incumbents. The response of indirectly affected voters indicates strong ‘demonstration effects’. I then discuss evidence that explains heterogeneity across the floods. Overall, the case allows to infer whether voters select prospectively or retrospectively.
How do policy positions of candidates affect vote choice under OLPR? Survey-experimental evidence using real candidates from Switzerland (with F. Quoß, and T. Däubler) [Abstract]
Some forms of list proportional representation, especially those with candidate voting, allow voters to express preferences in a flexible way. This is particularly important when the policy space is multidimensional or citizens have a preference for descriptive representation by candidates resembling them. However, it is questionable whether voters have in practice sufficient levels of information, in particular about candidates‘ policy positions, that would allow them to make the best-possible use of such systems. We examine how much information voters have about parties and candidates, whether they consider any limits to influence of candidates within parties, and how preferences for descriptive and substantive representation go together. We use a survey-embedded experiment that emulates the actual vote choice using real candidates and parties, just weeks after the Swiss National Parliament election of 2019. In a 2×2-design, we provide survey respondents with ballots that vary by information on candidates‘ policy stance in zero, one or two policy dimensions. We focus on the two most salient policy fields in this election, left-right and environmental protection. Initial findings suggest there is considerable learning, both about party and about candidate positions. In addition, preferences for being represented by candidates like oneself become weaker when information about policy positions is provided. We find no evidence that candidates‘ positions are discounted in comparison with party positions. Our results inform the debate to which extent preference voting may facilitate policy representation by individual candidates.
Environmental change and migration aspirations: Evidence from Bangladesh (with V. Koubi, and J. Freihardt)
[Link to Manuscript]
Can individual MPs damage their party’s brand? Evidence of a public procurement corruption scandal in Germany (with A. Leininger)
How does information affect vote choice in open-list PR systems? Evidence from a survey experiment mimicking real-world elections in Switzerland (with T. Däubler and F. Quoß) (under review)
Work in progress (original survey data collection completed)
Economic Inequality, Immigration and Redistribution: A Survey Experiment in Germany (research-led teaching project, LMU Munich)
Economic Shocks and Environmental Preferences – Evidence from Switzerland (with S. Gomm) (under review)