Photo: Johan Wingborg

MiMac- develop new research methods for pledge research

Research project
Active research
Project period
2020 - 2023
Project owner
Department of Political Science

Short description

An effective democracy requires political parties to offer clear choices to voters during election campaigns. However, as parties’ communication with voters has become increasingly fragmented and targeted, it is much harder for citizens to keep track of what parties promise. This threatens the quality of democratic representation. It also challenges established research methods for studying parties’ campaign promises. This project will develop new methods for studying parties’ promises in modern election campaigns. The project will integrate existing qualitative methods and develop new research tools based on the Artificial Intelligence subfields of machine learning and natural language processing. The project includes established experts on parties’ campaign promises, new media, machine learning and natural language processing.


This project aims to make fundamental advances in qualitative and quantitative methods for analyzing the promises that political parties’ make to citizens during election campaigns. It will establish Sweden as the innovative hub of the world’s largest comparative research program devoted to the study of campaign promises: the Comparative Party Pledges Project. The project will develop methods that will enable social scientists to study modern technologically advanced election campaigns. These new methods will also enable citizens to more effectively evaluate political parties’ campaign appeals. The methodological innovations we propose are highly relevant to a broad range of social science applications, which we will also demonstrate in this project.


Political parties’ campaign promises are a key part of democratic representation. Political parties play a critical role in linking citizens’ preferences and governments’ policies by offering voters choices in the form of promises during election campaigns, and then fulfilling those promises if they enter government after elections. Parties’ campaign promises are central to the mandate theory of democracy. Some scholars reject the idea of a mandate, while recognizing that parties’ campaign promises are a vital part of the democratic process.

Campaign promises are prominent parts of democratic practice in Sweden and other countries. Recent Swedish parties’ manifestos included promises “to reduce waiting times in emergency rooms”, “to increase the tax on alcohol”, “to raise the penalty for contributing to oil leaks”, and “to cut taxes for pensioners”. Specific pledges are made by other parties around the world. Comparative research demonstrates that parties’ election pledges refer to substantively important promises and are linked consistently with their more general ideological positions.

New technologies for fighting election campaigns are transforming the way in which voters receive information. Political micro-targeting allows campaigners to contact voters with information that is tailored to those voters’ interests, preferences and fears. Therefore, it is more difficult for citizens keep track of what parties are actually promising.

Research questions and method

This project will address the following substantive research questions in this field of research, as well as developing new methods for addressing these questions:

  • To what extent do parties make specific and distinct promises to voters during election campaigns?
  • How do parties’ specific election pledges relate to more general themes, issues and principles that distinguish parties from each other?
  • How are parties’ promises communicated to voters during election campaigns?

The abovementioned technological changes are also challenging established research methods for studying parties’ election appeals to voters. The sheer quantity of relevant information in modern election campaigns cannot feasibly be analyzed comprehensively with traditional qualitative methods based on human coding of documents. At the same time, new quantitative methods based on automated coding have been developed in isolation from established qualitative methods and related substantive research questions.

This raises the following specific methodological research objectives, which this project will pursue:

  • Develop new methods that automate the identification and comparison of parties’ promises and the ways in which they are presented during election campaigns.
  • Develop new methods that automate the classification of units of text and speech into policy themes and issues.

The new methods should produce results that are similar to and enrich those produced by human coding of the same information.



Research group

  • Elin Naurin, Professor of Political Science, University of Gothenburg, Principal Investigator
  • Robert Thomson, Professor of Political Science, Monash University, Principal Investigator
  • Christian Arnold, Senior Lecturer in Politics,  Cardiff University
  • Hannah Bechara, Postdoctoral Research Fellow, Hertie School of Governance
  • Rory Costello, Senior Lecturer in Politics, University of Limerick
  • Tom Dobber, Postdoc, University of Amsterdam
  • Tommaso Fornaciari, Research Fellow, Bocconi University
  • Dirk Hovy, Associate Professor of Computer Science, Bocconi University
  • Slava Jankin Mikhaylov, Professor of Data Science and Public Policy, Hertie School of Governance
  • Julia Runeson, Research assistant, University of Gothenburg
  • Claes de Vreese, Professor of Political Communication, University of Amsterdam
  • Lena Wängnerud, Professor of Political Science, University of Gothenburg.