Insurance Voting in the Centre: An Experimental Approach

Abstract

Abstract Recent research suggests that to help their preferred coalition win an election, voters are willing to vote for a political party other than their preferred choice. In this field, voting for smaller parties under proportional representation is an under-studied feature. A crucial factor to estimate the chances for smaller parties is polls. In this study, we analyze the influence of opinion polls on switching vote choice to a smaller party when the party polls are at different levels. Building on an original survey experiment, we elaborate the potential differences in impact on insurance voting for a small party with looser or stronger association with a government alternative. The focus is the 2022 Swedish general election and the three smallest parties in parliament: the Green Party (center-left), the Christian Democrats (right), and the Liberals (center-right). The experiment had nine different conditions where each of these parties was placed at different levels of opinion: below, at, and above the parliamentary threshold, while holding all other factors constant. We find that poll-induced insurance voting is most prevalent for the party with the strongest preference for a government alternative (the Christian Democrats) and least prevalent for the party with a more issue-focused stance (the Greens).