Electoral Systems and Democratic Legitimacy: lessons from the British and French Elections

In the past weeks, both the United Kingdom and France held Parliamentary elections. These countries have majoritarian-leaning electoral systems, which aim at ensuring the ‘governability’ of the country, and stand in stark contrast with the proportional systems characterizing countries like Germany, Belgium, or the Netherlands. 

These countries often require very long periods to form governing coalitions, which are forged through complex inter-party compromises that at times fail to make true to the promises taken during the electoral campaigns.

Some observers (for instance, Guerrera 2024) have therefore often praised the capacity of systems like the British ‘First past the post’ (FPTP) to deliver stable governments, as opposed to systems using proportional representation (PR). Yet, there are lessons to be learned by comparing the most recent elections in France and the UK. 

The French electoral system is characterized by two-turn constituency elections where a number of short-listed candidates (which can be more than two) qualify for a second round should no one reach more than 50% of the votes in the first round. 

The British system, conversely, assigns a given constituency seat to whoever wins the plurality (i.e., the relative majority) in a given constituency, regardless of whether or not the candidate reaches 50%. The results of these two systems are quite different. 

The FPTP system is notorious for delivering majorities which are often disconnected by the actual vote: for instance, in the most recent elections, the Labour Party won over 400 seats over 650 with less than 40% of the votes, the Liberal Democrats got 70 seats with 12%, but the far-right Reform party obtained only 4 seats, even though they obtained 14% of the votes. Differently, the two turns of the French elections had substantially different results. 

The first-turn results of the Parliamentary elections were relatively close to the results of the European elections held a month before under proportional representation. 

Given the high turnout, a large number of constituencies resulted in second-turn races with three candidates; the results of these races were ultimately determined by agreements between two of the three candidates in each constituency, forged to to avoid splitting the vote against the far-right Rassemblement National (RN). 

Ultimately, the second round of the elections saw the RN, which had triumphed in the first round and in the European elections, being beaten into the third position by such tactical voting and tactical withdrawals.

Comparing these elections yields important insights on the democratic legitimacy of electoral systems. To understand those, we need to set two assumptions. 

First off, voting and political preferences are ‘complex’. They are both ‘positive’ (appreciation for a candidate), but also ‘negative’ (opposition to a candidate). 

Second, preferences are ‘layered’ rather than univocal. Most voters will prefer, say, candidate A over candidate B & C, but candidate C over candidate B. Their ‘political utility’ is maximized when candidate A wins, but if candidate B wins they are worse off than if candidate C had won.

Input Legitimacy

We need to define the criterion to assess the legitimacy of a democratic system. Typically, political legitimacy stems from ‘input’ (the congruence of the outcomes of a voting system with political preferences over candidates), and ‘output’ (the congruence of the outcomes of policy-making with the desired policy outcomes). 

Output legitimacy can be provided by many political systems, not only democracies; democracies should shine, instead, in providing ‘input’ legitimacy. 

To do so, democracies must maximize the congruence of the electoral results with the preferences of the citizens. In other words, the ‘aggregate voter utility’ stemming from the congruence of the results with preferences is our metric of the extent to which a system has input legitimacy, and it is therefore democratic. 

Typically, as argued for instance by the German Federal Constitutional Court (2009) proportional representation is seen as fulfilling better the principle of electoral equality, and therefore more effective in delivering input legitimacy than systems like the British FPTP, which trades some electoral equality (and therefore input legitimacy) for more stable governments (and therefore more output legitimacy).

Both systems, however, fail to account that preferences are both layered and bi-directional (‘for’ some candidates, but also ‘against’ others). 

The French system, instead, is better placed than both the British FPTP and classical PR in maximizing congruence with complex preferences. Since preferences are layered and bi-directional, asking people to reallocate their first-turn preferences, among candidates in a shortlist is more congruent with 'complex' individual preferences than a one-shot vote that assigns a seat to whoever wins the plurality (like in the FPTP). 

Both proportional representation and FPTP allow voters to target the first part of their preferences or the second, but not both: voters are asked to decide whether to vote for the candidate they prefer or against the candidate they dislike, but not necessarily these preferences imply voting for the same party. 

The French system, instead, allows voters to focus their first-turn preference on their most preferred candidate, and then decide, in their second-turn preference, whether to continue voting for their preferred candidate, or vote against a disliked candidate instead.

Relatedly, voters do not have simple, univocal preferences but complex, layered preferences, and the French system allows voters to express these preferences in ways neither the FPTP nor PR can achieve. Take the case, for example, of a voter whose preferred candidate did not make it to a second turn.

If you assume that that voter had truly univocal preferences for the excluded candidate, then she would abstain at the second turn. In turn, the more voters hold truly univocal preferences the more the outcomes of the French electoral system mirror FPTP. 

But in reality, most voters hold complex preferences, and the fact that results are so different demonstrates that preferences are indeed layered. As a result, the aggregate voter utility (our proxy for how democratically legitimate a system is, if democratic means representative of people's will) is higher in this system than in both FPTP and PR.

Of course, there are systems that might be highly democratic but inefficient, resulting in paralyzed parliaments unable to govern, ultimately undermining output legitimacy. 

Input legitimacy is not the only thing citizens care about, and perhaps not even the most important one for some: outcomes may matter as much or even more than process. 

In other words, electoral, input legitimacy is not the only thing a political system must maximize; the capacity of the government to deliver on citizens' expectations (its output legitimacy) is as important. 

However, there are other arrangements that can boost governmental stability. Taking another page from the French constitutional playbook, one could be the principle of reverse confidence, whereby it is up to the parliamentary opposition to muster enough votes to trigger a non-confidence motion in the government (and therefore, some parties might abstain if they do not participate in the government coalition but do not intend to replace it or trigger new elections). 

Government performance can further be improved by a higher involvement of experts in policy-making and more rigorous non-partisan oversight, and furthermore, even systems like FPTP can result in ‘hung parliaments’, very unstable governments (for instance, the British Conservatives changed three prime ministers in four years despite the 200 seats majority) or disappointing policy outcomes. 

Non-democratic systems can also provide effective policy outcomes; what makes democracies stand apart is the capacity to affect the outcomes by better tailoring the input to the people’s expressed preferences, and – as argued in this column- double-turn constituency elections are more effective in doing so than both FPTP and RP, and should be considered by countries seeking to improve their electoral processes and input legitimacy. 

Conversely, other (non-electoral) institutional arrangements can be introduced to ensure governmental stability and better policy outcomes, improving government responsiveness and output legitimacy.

IEP@BU does not express opinions of its own. The opinions expressed in this publication are those of the authors. Any errors or omissions are the responsibility of the authors.

If you want to stay up-to-date with the initiative of the Institute for European Policymaking@Bocconi University, subscribe to our monthly NEWSLETTER here.