Prospective vs Retrospective Voting: Reflections from Two Norwegian Surveys

While Norwegian citizens believe voters cast their vote on the basis of both previous actions and the political plans for the future, a majority of Norwegian representatives consider voting to be retrospective only.

By Andrea Fumarola*

A few weeks ago, almost 3 million Norwegians voted in the local elections to renew 356 municipal and 10 county councils throughout the country. While the four governing parties at the national level lost votes in comparison with four years ago, the main opposition party, the Labour Party (Ap) couldn’t take advantage of it, losing support all across the country. The ‘winners’ of these elections were three parties: the pro-farming Centre Party (Sp), the Green Party (MDG) and the new protest party against road tolls (FNB) that got 16.7 percent of the votes in Bergen. It is difficult to analyse these results from a ‘national perspective’ and assess if (and how) they will reflect upon both the governing and opposition parties. Nevertheless, this election gives room to make some general reflections on dynamics of voting behaviour, but also to narrow the scope to the Norwegian case analysing specific survey data.

In representative democracies citizens ideally hold the government accountable for its actions. With their power to punish or reward parties at elections, voters express their preferences and force political elites to perform in line with these. Most of the empirical research on electoral accountability states that voters judge parties ‘retrospectively’: rewarding or punishing the government for good or bad performance in several policy (e.g. economy) as well as non-policy (e.g. corruption) fields. The central idea is that voters stand in a principal-agent relation to parties – a vertical linkage – and that they evaluate their agents by performance ex post. This mechanism represents the essence of the retrospective voting theory.

However, this is not the only approach adopted to study electoral accountability. It has been confronted, with another fortunate approach originally elaborated in the 1950s, one that adopts a ‘prospective’ vision of voting. The source of conflict between these approaches lies in the way in which voters cast their vote. According to the so-called Downsian approach, citizens vote for a party they believe will provide them with higher utility than any other party during the electoral period. In other words, voters make a choice comparing future performance they expect from the competing parties.

Thanks to survey data from the Norwegian Citizen Panel (NCP), we could investigate citizens’ view of elections as means to hold governments accountable in the Norwegian context. Respondents were asked to assess the way people vote in the national elections, i.e. whether they judge politicians’ previous actions or, rather, look at their future political plans. The picture obtained is rather interesting.

Figure 1. Citizens’ assessment of voting behavior

Fumarola Fig 1

Source: Norwegian Citizen Panel (NCP) round 12, 2018. N = 1349

Quite surprisingly, Norwegian citizens consider the evaluation of parties’ previous actions as important as their electoral pledges for deciding how to vote. Half of the respondents, weight these two elements equally when formulating their preference for a specific party. Only 28 percent of respondents evaluates previous actions as crucial for their electoral choice. Given the findings in contrast with the main theory of voting behavior, we investigated whether other factors like the propensity to vote for a specific party, belonging to the governing or opposition front, influenced responses to this question. However, we did not find any significant pattern among party supporters of the main national parties. Voters show a rather stable distribution of preferences. Resembling what was found by Michael Lewis-Beck in his comparative study on six Western countries, our data suggest that Norwegian voters evaluate parties and candidates simply weighting “what they say that they will do, as well as what they have done”.

However, the same question has also been included in the Panel of Elected Representatives (PER), a survey annually fielded to analyze the opinion of Norwegian representatives elected at the municipal, county and parliamentary level. The unique possibility to field questions in parallel to the Citizen Panel allow researchers to compare responses at the supply and demand sides of political representation. In fact, given the distribution of responses registered at the citizen level, the data presented in Figure 2 are quite interesting.

Figure 2. Representatives’ assessment of voting behavior

Fumarola Fig 2Source: Panel of Elected Representatives (PER) round 1, 2018. N = 4165

When compared to their voters, elected representatives consider voters more concerned with past performance. For a consistent majority of Norwegian politicians, people rely almost exclusively on the previous actions of parties when they cast a vote. This ‘backward’ vision of elections could have important implications for the broader representation process. In her seminal article Jane Mansbridge lists several potential models of representation involving different dynamics between voters and representatives. Among them, and strictly connected with the retrospective voting theory, there is the concept of anticipatory representation. According to this view, representatives perceiving citizens more focused on their past performance will “try to please future voters” in an attempt to “anticipate (their electoral) reactions”. In this way the traditional principal-agent relation goes backward, with representatives pushed to adapt their policies to please voters in the next election, instead of sticking with their election pledges. The reasons driving these different views of voting behavior are rather than clear.

A potential explanation could be found in the parliamentary and pure proportional nature of the Norwegian political system. Recent studies show that democracies governed by single-party coalitions – such as UK, Spain or Canada – present the highest percentages of pledge fulfillment, while countries with more fragmented party systems and governed by coalitions of parties show the lowest percentages. These dynamics seems to vary across political systems in ways that reflect power-sharing arrangements. Norway is a parliamentary democracy frequently governed by coalition or minority governments as the last four Solberg cabinets prove. In such a context, parties will inevitably find much more difficulty fulfilling their promises, being forced to change and adapt their agenda to the competing interests of the coalition partners. The government agenda will therefore follow lines of policy that are often different from those promoted during the electoral campaign in their platforms and rallies. The data from NCP and PER surveys seem to support the idea that the perceived impossibility to fulfil pledges pushes representatives to consider performance more important for voters’ choice.

This look at Norwegian citizens’ and representatives’ opinion on a specific aspect of political representation has shown the connection between theories of voting and the broader concept of political representation. What emerges from these data is the divergent opinion between the two sides of the chain of representation that we have tried to explain with reference to the characteristics of the Norwegian political system. While citizens see prospective and retrospective evaluations having about the same impact on voting behavior, a consistent majority of representatives consider voting to be retrospective. Several ongoing projects within The Politics of Inequality Project using these data focus on the correspondence of preferences and views between voters and representatives to understand and explain what factors enhance congruence and what elements drive unequal representation.


* Andrea Fumarola is postdoctoral researcher at the Department of Comparative Politics, University of Bergen


Legg igjen en kommentar

Fyll inn i feltene under, eller klikk på et ikon for å logge inn:

Du kommenterer med bruk av din konto. Logg ut /  Endre )


Du kommenterer med bruk av din Google konto. Logg ut /  Endre )


Du kommenterer med bruk av din Twitter konto. Logg ut /  Endre )


Du kommenterer med bruk av din Facebook konto. Logg ut /  Endre )

Kobler til %s