Introduction

THEORIES OF REPRESENTATIVE DEMOCRACY ARE RELATIONAL, FOCUSING on connections between voters and elected representatives. Naturally, representation has to do with relationships between what economists call principals (voters) and agents (elected representatives). Agents are supposed to act on behalf of the principals, like decide on the laws regulating the working of society or set tax levels. Principals have one essential job – electing the agents. Eventually principals also give agents instructions/mandates and try to ensure that agents do their job by retrospectively holding them accountable.

Representational relationships can be studied statically emphasizing differences between levels, between principals and agents. Hanna Pitkin talks about standing-for representation. How large should differences be between principals/the people and agents/elected representatives? Do agents have to be similar to principals in order for representation to function well? Many studies of social representation and policy representation have analyzed the problem. As a matter of fact, empirical studies of political representation have been dominated by these kinds of static level comparisons. That goes as well for the limited number of comparative projects that have been done on political representation (Miller et al. 1999; Schmitt and Thomassen 1999; Esaiasson and Heidar 2000).

The reason for this focus on static level comparisons has not been theoretical. Students in the field have on the contrary always been conscious of the fact that representative democracy harbors an interplay component between principals/voters and agents/legislators. Instead, the main reason for the dominance of static level comparisons has to do with a lack of good time-series data dealing with the relationship between principals and agents. This is especially true in political science where it so far has been very costly to collect comparative data on voters and legislators; particularly if one wants to study the opinions of voters and elected representatives. Surveys are expensive and they have a very noticeable access problem on the elite level. Elected representatives are often reluctant to participate in surveys. Response rates are very low in most elite surveys making all inferences tenuous (Brothén and Holmberg 2009).

The central issue in representative democracy theory has to do with power. From a normative standpoint it is the principals/the people who should have the power – they should decide what the agents/the representatives do. The will of the people shall rule.