A global analysis of how losing an election affects voter satisfaction with democracy

How easily do people turn against democracy? In a recent paper, we asked whether losing one election is enough to sour voters on the whole idea of democracy. We find that indeed it can be – if their democracy is relatively new.

In addition to this difference between established and emerging democracies, we also find another important pattern: among the established democracies, the type of electoral system affects loser satisfaction, but in newer democracies, it matters much less.

These two findings suggest that we need to considerably expand our understanding of “loser’s consent”.

In the aftermath of an election that hasn’t gone their way, voters can cast blame in any number of directions. In our paper “A global analysis of how losing an election affects voter satisfaction with democracy”, recently published at the International Political Science Review, we ask if voters will blame the entire democratic system.

Using data from public opinion surveys around the world (the AfroBarometer, AsiaBarometer, EuroBarometer, and LatinoBarómetro), we looked at the relationship between supporting a party that had lost the last election, and supporting democracy overall in that country.

We found, unsurprisingly, that “losers” were generally less supportive of democracy than were “winners”. But the more interesting findings are about the factors that affect losers’ satisfaction with democracy. In some countries, loser satisfaction is relatively high, whereas elsewhere it is relatively low. So what explains these cross-national differences in losers’ satisfaction?

Prior explanations of these cross-national differences have focused on the role of electoral rules. Specifically, majoritarian electoral rules give no reward to parties that finish in second place, whereas proportional electoral rules assign seats to large parties even if they weren’t the largest. So the winner-loser gap in satisfaction should be smaller in proportional systems. After controlling as best we can for all other sources of cross-national variation, we find support for this theory – but only in Western Europe. In newer democracies outside Western Europe, electoral rules do not have this effect.

Figure 1 below demonstrates this. It has ‘democracy years’ on the x-axis, so more established democracies are towards the right-hand side and newer democracies are on the left. The y-axis is the size of the coefficient on our ‘electoral system’ variable, which we code to increase with the proportionality of the rules. So if proportional systems lead to more satisfaction, values on the y-axis will be high. We show two sets of coefficient values, one for supporters of the ruling parties or “winners” and one for supporters of the opposition parties or “losers”.

Figure 1: The Effect of Electoral Rules on Winner-Loser Satisfaction With Democracy, In New And Old Democracies Around the World

As Figure 1 makes clear, proportional electoral rules are associated with more satisfied losers, but only in countries that have been stable democracies for some time – the cutoff for statistical significance is roughly 40 years.

Electoral institutions are clearly important, but our findings reinforce the insights of other authors that it is unclear how such institutions perform (or even arise) when risks are high and stability is low.

Satisfaction with democracy seems to work very differently in different areas of the world, and we should not assume that theories developed in Western Europe will necessarily travel well – either temporally or geographically.

We suggest that future work might find it useful to re-conceptualise losers’ consent, accounting for more of this global variation. One fruitful avenue for research might to split the concept into attitudinal and behavioural components. That is, perhaps in Western Europe, people take democracy for granted, and although they may report attitudes of dis-satisfaction, they are a long way from turning those attitudes into actual protest behaviours. If democracy is newer, the attitude-behaviour gap may be smaller, affecting the dynamics of loser satisfaction in important ways. In a paper that will be presented at APSA later this year, we present research moving in that direction.

This blog post is based on research by Ben Farrer and Josh Zingher