Strategy is key in assessing the proportionality of the Single Transferable Vote

The Single Transferrable Vote is in use in several countries, most notably Australia, Ireland, India, and Scotland. It is held up as one of the more proportional systems, and frequently advocated as an alternative to the First Past the Post system in use in the UK. But just how proportional is it? David Farrell argues that determining the answer rests on understanding party strategy.

Malta, where STV is in use (Credit: John Haslam, CC BY 2.0)

The single transferable vote (STV) is an unusual electoral system. It is popular with reformers and many electoral systems specialists yet remains rarely used for national level elections (the only three cases being Australia, Ireland and Malta). Despite its rarity it is prone to considerable variation, leading some to question whether it is in fact a single electoral system but rather ‘an evolving family of vote-counting rules’. And it has a proportionality record that is poor by comparison with other electoral systems to such an extent that it is often referred to as a ‘semi-proportional’ system.

In an article in a recent issue of Representation Richard Katz and I assess the proportionality of STV systems in the light of the fact that the species varies from one arena to the next and that this can inter-play with how parties and voters operate strategically.

STV strategies

Strategic calculations reveal themselves in a number of areas, most particularly: decisions by parties over how many candidates to field in an election; vote-management tactics over how to deal with preferences at the end of a party’s run of candidates (i.e., inter-party transfer flows); and strategic calculations by voters about how to use their transfers.

The three cases of nationwide STV elections reveal interesting strategic variations in all of these areas. Starting with candidate selection, we see that Irish parties seek to maximize the efficiency of their vote by minimizing their number of candidates – the larger parties running roughly as many candidates as the number of seats they expect to fill, and the smaller parties running just one candidate, sometimes two. In Australia parties feel no such need to limit the number of candidates they run. This is due to rules that require voters to express at least as many preferences as there are seats to fill coupled with the role that ticket voting plays in most of the Australian upper house cases (more generally, see Farrell and McAllister 2006). As a result parties tend to run as many candidates as there are seats in the constituency.

In Malta, the count back system used to fill vacant seats – which enables parties to field ‘star’ candidates to attract large numbers of first preferences before stepping aside – combined with a rule that ‘reinforces’ the seat gains for the party with the most first preference votes in an election results in a strategy to field many more candidates than there are seats to fill.

Similarly there are large strategic variations in how parties promote inter-party preference transfers: Maltese parties actively discourage them; Australia’s ticket-vote results in full transfers across parties; Irish parties tend to be somewhere in-between the two.

And then, of course, there is the issue of how the voters use their right to transfer preferences from one candidate to the next – in part influenced by systemic rules, such as minimum preference requirements that Australian STV systems tend to hoist on their voters – which (as we demonstrate) can have significant impacts on party election outcomes.

The impact of party and voter strategy on STV proportionality

In our analysis we plug these strategic calculations into a multivariate model to examine how they influence proportionality outcomes in post-war elections in Australia (Senate elections), Malta and Ireland, finding a number of interesting patterns, most notably:

Nomination strategies perhaps have the greatest impact, with some evidence that (counter-intuitively) the Maltese practice of over-nominating might actually have a positive impact on proportionality;
Another Maltese practice – that of promoting inter-party ‘exclusivity’ in transfers – also appears to have a positive impact on proportionality;
The Australian ‘vote above the line’ (or ‘ticket voting’) impacts negatively on proportionality;
The impact of district magnitude on proportionality is quite slight, and certainly less than that of strategic calculations

—

Note: this post represents the views of the author and not those of Democratic Audit or the LSE. Please read our comments policy before posting.

—

David Farrell is Professor of Politics at the University College Dublin. He is also the President of the Political Studies Association of Ireland