STV & STV+

STV was the recommendation of the 2004 BC Citizens’ Assembly on Electoral Reform. Adjacent ridings are merged to form multi-member districts, and voters use a ranked ballot to express their preferences between individual candidates. STV+ adds a small top-up layer to improve proportionality.

In 2004, the BC Citizens’ Assembly on Electoral Reform recommended that we adopt the Single Transferable Vote (STV) because they felt it best matched their top values:  (1) proportional representation, (2) effective local representation, and (3) voter choice.

What is STV?

STV was initially invented over 100 years ago and is the most widely used form of proportional representation (PR) amongst the Westminster democracies;  it is currently used nationally in the lower houses of Ireland and Malta, the Senates of Australia and India, in the devolved assembly in Northern Ireland, and at the state and/or municipal levels in Australia, New Zealand, Scotland and two cities in the USA (Cambridge and Minneapolis).  At one point, it was used in over twenty cities in western Canada (at both the municipal and provincial levels) and over twenty cities in the USA (including New York City, Cincinnati and elsewhere).  For the most part, STV was repealed by the governments of the day because it was effective in giving opposition voices a seat at the table.  For example, under STV, African Americans and socialists were elected to city councils in the USA when they couldn’t win seats in first-past-the-post-type elections.

Percentage of women elected in the Australian Senate (STV) and House of Representatives (AV)

In the Australian Senate, the representation of women increased much more rapidly under STV than under their majoritarian lower house system (Alternative Vote), reaching nearly 40% in the last two elections, as compared with only 25% in the lower house.  In Ireland, 55% of the 2014 European Parliament contingent (elected using STV) were women. The Australian Capital Territory currently has 52% women elected using 5-seat STV, and Tasmania has 44% women. The Icelandic Constitutional Assembly elected 40% women in 2010.

For a more complete discussion about STV, please visit our companion website, STVforCanada (note that although FVBC does not explicitly endorse one voting system over another, we regard STV as a first-rate option for BC and Canada).

How STV Works

STV merges existing single-MP ridings with one or several of their neighbouring ridings to create local regions covering anywhere from 2 to perhaps 7 or 10 current ridings (e.g., the metro Victoria region currently has four MPs, so this would be a natural size for an STV district there – see a detailed example, including a map, at STVforCanada;  the city of Vancouver currently has six seats, which might be the largest natural district in BC, and the southern interior has six seats which might most reasonably be grouped into two sets of three ridings).  The ultimate choice of electoral boundaries would be the responsibility of a non-partisan Electoral Boundaries Commission, just as it is with our current system.

The graphic below illustrates the key ideas behind STV:

Overview of Single Transferable Voting (courtesy of Library of Parliament). Note that even simpler counting rules could be used; in particular, it is not necessary to define a quota – votes can simply be transferred in rounds from the candidate with the fewest votes in each round until there are only as many candidates left as seats available. This approach has been used in some South Australia towns.

And here’s a typical ballot for a four-seat STV riding:

STV Ballot - Ontario
Typical STV ballot for a hypothetical four seat region in Ontario

Within each STV multi-member riding (district), one or more candidates from each political party, as well as independents, would be nominated for the ballot.   Voters would then rank one or more of the candidates in the order of their preference.  STV is widely recognized as offering voters maximum choice, allowing them to choose not only candidates within a party, but across parties as well.

The candidates with the greatest support would then be elected through a counting process in which the candidates with the least support are sequentially eliminated and the ballots transferred to each voter’s next most-preferred choice such that the final group of candidates elected broadly represents the diversity within each electoral district.

National Results Using STV

We simulated the effect of using STV in the last federal election.  Recognizing that we had to make a number of assumptions for this exercise, including the assumption that voter behaviour wouldn’t change significantly (which it likely would if we were using a voting system that encouraged more honest voting), we estimate that, had voters voted roughly as they did in 2015, then, as shown at STVforCanada, the Liberal Party would likely have won about 43% of the seats on 40% of the vote (instead of the 54% they won under our current system), the Conservatives would have won about 33% of the seats on 32% of the vote, the NDP 19% of the seats on 20% of the vote (instead of only 13%), the Bloc 4% on 5%, and the Greens <1% on 3%.  Overall, the results would have much more accurately reflected how voters actually cast their votes, and we would have largely eliminated the regional imbalances – there would have been Conservative and NDP MPs elected in Atlantic Canada, Liberal MPs elected across the Prairies and in the BC Interior, and Conservatives elected on Vancouver Island.  And even though the Liberal Party might have won a few more seats than their share of the vote might have suggested, they would not have won nearly such a large bonus and would have to collaborate with one or more of the other parties in order to pass legislation.

Making STV More Proportional

As noted above, the Green Party would not likely have won more seats under the basic form of STV described above. This is because their supporters are very diffusely distributed, rarely representing more than 5% of the voters in any given region. Since STV districts are usually made up of only a few seats, Green Party supporters can only elect a Green MP in areas where they constitute about one seat’s worth of voters. For this to happen, there has to be some mechanism that would allow them to consolidate their votes from over a broader region. Fortunately, this is quite easily done – in principle, all we have to do is to add a top-up seat to a suitably large region.

For example, we might consider adding roughly one top-up seat for every 7-10 regular STV seats. If we did this in the BC Interior, which currently has 9 federal seats, we could imagine having two three-seat districts in the southern interior, one two-seater around Prince George, and perhaps a single seater in the northwest. We could then add one regional seat to cover the entire interior and elect a candidate to the regional seat to reduce any imbalance arising from the district elections. If Green Party supporters were numerous enough (say about 10% of all voters), then they could elect a regional Green MP; if not, then the top-up MP would likely represent supporters of another party (or possibly an independent) that did not win as much representation as they deserved in the regular districts.

We simulated how STV+ might have changed the results in the 2015 election. The biggest change would have been that Green Party supporters would have elected 8 MPs instead of just one, so it would have been effective in enabling these voters to elect their preferred candidates. It would have achieved this by slightly reducing the number of Liberal and Conservative MPs elected – by about 1% each – which would have brought both the seat shares of these parties slightly closer to their actual vote shares, so the overall result would have been fairer. We therefore feel that STV+ would be an excellent choice for Canada to adopt.