Broken promises?
Do political parties deliver what they promise or are their words just empty platitudes for the purpose of winning elections? – A question that is playing a significant role in the local and European elections in June this year, but also in Elisa Deiss-Helbig’s research. Does this mean that politicians neglect certain population groups and treat them unequally? – This is what the political scientist wants to find out.
Through her research, Elisa Deiss-Helbig wants to contribute to the debate on the ubiquitous topic of political inequality. Her project deals with questions such as: “Which social groups are given a lot of election promises, which tend to be given fewer or none?”, “For whom are election promises actually implemented?”, “How can we explain these discrepancies?”, “For which groups does the population / do specific population groups support election promises and their implementation?”.
“‘Inequality’ is not only a topic that plays a central role in academic debate but also an omnipresent topic in public discourse and of great socio-political importance, especially with regard to social cohesion,” says Elisa Deiss-Helbig, explaining why she chose her research topic. “There are many forms of inequality. Central to my research are political inequalities, in particular inequalities between social groups with relation to the promise and receipt of advantageous political measures (that is, policies) or goals (that is, outcomes). I am not limiting myself here to individual social groups (e.g. women, income brackets, age groups...). Instead, I am examining these inequalities for all relevant social groups. However, there can also be political inequalities vis-à-vis the presence of certain groups in politics (in parties, parliaments, but also in terms of political participation). Then there are, of course, also social inequalities, which are not, however, part of the direct object of my investigation; these are inequalities in terms of resources and living conditions. However, political and social inequalities are closely interwoven and cannot be considered completely separately.”
German-French cooperation funded by the German Research Foundation (DFG) and the French National Research Agency (ANR)
Together with Theres Matthieß from Trier University and Isabelle Guinaudeau from Sciences Po Paris, Elisa Deiss-Helbig is leading a collaborative German-French project that differs from previous studies in three respects: Firstly, from studies that tackle the reasons for political inequality. “Our project differs here in that we start investigating inequalities at a relatively early stage of the political process, namely with election promises. Most studies in this area look at political measures (policies) that have actually been implemented.” Secondly, the project differs from studies on election promises and their implementation insofar as its central focus in explaining the likelihood of implementation lies on the characteristics of promises, and in particular the target populations of the promises. Previous studies, she says, have primarily made more comprehensive country comparisons and concentrated on the influence of institutional factors (system and form of government, economic situation, etc.), i.e. on the macro level. And finally, their project is especially distinguished by the fact that Elisa Deiss-Helbig and her team have chosen a very comprehensive research design and are shedding light on factors on the supply side (parties and their election promises) as well as the demand side (voters).
Elisa Deiss-Helbig
German politics
Experimental research and qualitative case studies for developing innovative ideas
To be able to illuminate and compare the parties and their election promises on the one hand and the views of voters on the other, the political scientist has developed a very comprehensive research design. “To explore our questions, we opted for a mix of methods that takes the diversity of the various work packages into account in the best way,” explains Elisa Deiss-Helbig. The variety of methods and approaches is definitely a big plus, she adds, but also calls for wide-ranging experience with and knowledge of sometimes very different methods. That is why she describes working in a team – where everyone can contribute their special knowledge and complement each other very well – as a great advantage. “My research is mainly quantitative, but we also use qualitative research methods.”
Analysis of election manifestos, press reports and people’s views
“Specifically, we conducted a quantitative analysis of the content of election manifestos to collate the election promises of the German governing parties and French presidents over a period of about 20 years and then manually coded the target groups of the promises by means of inductive logic,” the political scientist continues. ‘Inductive’ means not searching for predefined groups in the texts but documenting the groups as they appear in them. This ensures that no group is overlooked. The many individual groups are then subsumed under larger group categories.
“At the present time, we are looking at press reports or announcements by the government, parties or ministries to check whether promises have been implemented,” says Deiss-Helbig. These data are then evaluated using quantitative statistical methods. “Furthermore, we are currently collecting longitudinal data – that is, over the last 20 years at various points in time – and cross-sectional data – that is, for Germany and France – on various characteristics of the target groups identified in the promises.” Among these characteristics are above all voter turnout in these groups and the image that these groups have within society. The latter is determined on the basis of an automatic analysis of media data using machine learning methods.
As the political scientist is also interested in the population’s views, she has conducted three online surveys to determine its support for election promises – two in Germany and one in France; further surveys are planned for the end of 2024. The aim of these surveys is to ask, in the classic way, about people’s attitudes towards various social groups, political issues and their voting behaviour, but they also include survey experiments. “We showed respondents fictitious campaign posters with election promises, for example, and told them about the implementation of these promises. The aim was for us to be able to draw causal conclusions, in this case from “group appeals”, that is, appeals to certain groups in election promises to support the governing parties, for example.”
Aligning the methods
The research team conducted the first study in Germany in the autumn of 2021 and in France in the spring of 2022. After analysing the data, two particularly critical points came to light: “We had attached a lot of importance to external validity in the survey experiments,” explains Elisa Deiss-Helbig, who conducted the survey during the German parliamentary and presidential election campaigns. In other words, the experiments were designed to be as realistic as possible, which is why the team drew on real promises from the election manifestos of the later governing parties. “We then informed the respondents that the government had fulfilled or broken the promise and examined whether this influenced the respondents’ satisfaction with the government, for example,” she reports. However, according to Deiss-Helbig, recoursing to real promises led to limitations in the choice of campaign promises, meaning that the researchers had to take great care in their analyses to ensure that potential problems of internal validity – i.e. the least possible influence of potential confounding variables in the experiment – did not affect their results. “We had to make sure, for example, that the respondents’ reaction was genuinely due to our treatment, that is, the status of implementation and the target group of the promise, and not, for example, to the specific measure that was promised.”
For this reason, the team decided to confront the respondents in the second round of the survey with hypothetical scenarios in order to eliminate, as far as possible, factors that could influence the experimental design. This was to be achieved by showing respondents fictitious campaign posters that promised more funds for measures of central importance to the social group addressed in the promise, such as child benefit for parents, pensions for pensioners, better infrastructure for the rural population.
“In the second round of the survey, we also systematically polled the subjective aspect of group affiliations for a large number of social groups. In the first round of the survey, we only did this for individual groups.” This allowed the team to determine how big the differences are between objective – measured primarily via sociodemographic and structural characteristics – and subjective group affiliations.
Initial findings
According to Elisa Deiss-Helbig, the collection of data on the election promises, especially the fulfilment of the promises, has not yet been completed. This means that she cannot report any consolidated results at the present time. “However, what we can already say is that the election promises address a large number of social groups and that these also include many groups described in research as marginalised,” says the political scientist, summing up her initial findings. “In addition, the groups addressed in Germany and France are similar. This preliminary result is by all means consistent with the assumption that a relatively wide range of different groups are addressed at the level of election promises, or pledge making, in order to mobilise as many groups as possible.”
As a preliminary result of the population surveys, the team found that respondents always reacted negatively when promises were broken. However, they could not identify any positive effects when promises were fulfilled, not even if the respondents ought in fact to benefit from the promise being fulfilled or if it is advantageous for groups considered to be entitled to political measures that benefit them. In this regard, Deiss-Helbig will conduct further analyses to check the robustness of these results.
Outlook
The current study will continue for almost two years. Beyond that, there are several aspects that Elisa Deiss-Helbig would like to pursue more closely in subsequent studies. “One aspect would definitely be the role of various overlapping characteristics of social groups in terms of how they are perceived both by the population and in election promises. Do certain groups in society generally have a rather negative image and, accordingly, is there little support for political measures that benefit this group, or does this image change (in a positive or negative sense) when certain (socio-demographic or socio-structural) characteristics are added? Are certain social groups only promised measures that benefit them in a specific configuration of characteristics and what role does this play in implementation? Or is it fundamentally difficult for some groups, regardless of the specific group configuration (literature, for example, speaks of an “immigrant penalty” that is hard to overcome)?
Participants are recruited by survey institutes entrusted with conducting the surveys – in most cases by means of panels. Quota samples are commissioned for Elisa Deiss-Helbig’s surveys, i.e. participants are selected according to quotas specified by the researchers (in this case: age, gender, region, level of education). The aim is to achieve the highest possible representativeness for the overall population (voting population). In the event that quotas are over- or under-fulfilled, there are weighting variables that can be used in the statistical analyses of the data.
The initial focus of Elisa Deiss-Helbig’s project is a German-French comparison. Although Germany and France are two countries that are similar in many respects (e.g. in terms of the level of democracy or prosperity), they demonstrate major institutional differences (particularly in relation to the political system and the party system). At the present time, Elisa Deiss-Helbig is contemplating an extension of her research work to include other countries, for example by cooperating with research projects that are examining similar questions.
In the context of the local and European elections on 9 June 2024, the University of Konstanz has published a press release and an article on the topic in “campus.kn – The online magazine of the University of Konstanz” under the heading “Science Backstage”:
Further information about Elisa Deiss-Helbig
polver.uni-konstanz.de/en/zuber/team/translate-to-englisch-elisa-deiss-helbig/