STV vote transfers democratic but not proportional based on first votes
Proportional representation means different things to different people.
Canadians who are asking for PR not asking for PR because they want each province to have representation based on population.
We already have that.
In Canada currently, seats are allocated as per population, with a minimum of four seats for each province and one seat for each territory.
This is Representation by Population*
And our election system is fair in that way,
In one view, one could say Rep by Pop. produces unequal representation in that each province or territory does not get the same number of seats. But as far as votes go, Rep by Pop. is fair - each voter has at least the potential to get the same representation. Of course, the inequities and inefficiencies of FPTP means that such fair and equal representation is pretty much blurred out of all recognition.
Our drive for PR in Canada does not arise from our need to have rep by pop., as the U.S. applauds itself for having, at least in the House of Reps., but it arises from desire to start with that as an already-existing base and add reforms that produce equal voting.
Some say that vote transfers under STV produce dis-proportional results.
==============================
So I would ask--
if NZ brings in a second choice (a back-up preference) for party vote system for its MMP elections, would you think a result that is produced by vote transfers under that system is dis-proportional if it varies from the way first-choice preferences are marked?
=====================
More generally, I want to point out that any system that uses transfers to avoid vote waste produces results that would vary from first choice votes.
This is another thing, like the respective advantages and disadvantages of party-list PR versus STV, where the values a person holds determines whether the person favours a system or not.
Each is good in its own way. But no system can be everything to everybody.
Adding a back-up preference to the usual FPTP component of MMP is a step toward better effectiveness of voting but does not go far enough – you should allow votes to rank as many parties as he or she wants.
But at the same time, make it voluntary. - I disagree with requiring voters to mark back-up preferences if they don't want or care to.
or do away with the second vote altogether and just go by the first vote and back-up preferences for that vote.
There is talk of people voting one party and electing the local member and expecting that, they had marked a different party for their party choice, so getting the benefit of two different votes.
One vote with back-up preferences - used only if first choice is not used - that would be used for both local and top-up seats - would prevent that possibility.
Any system that uses vote transfers allows the final result to not be proportional to first choice votes.
Some say that variation from the first choice votes is a dis-proportional results.
This would mean that STV is dis-proportional.
Interesting that some of the same parole who quibble about STV producing dis-proportional results see the waste of votes under the party vote (top-up) component of MMP and seek to address it in the same way that STV does — through back-up preferences and vote transfers.
So I would ask them
If you are able to bring in a system that uses a second choice (a back-up preference) for the party vote system in NZ, would you still think a result that is produced by vote transfer under that system is dis-proportional if it varies from the way first-choice preferences are marked?
Or would you learn to see the final result unde such a system fair even if it varies from the first-choice votes cast?
Any system that uses transfers to avoid vote waste would produce results that would vary from first choice votes.
This is another thing, like the respective advantages and disadvantages, of party-list PR versus STV, where the values a person holds determines whether the person favours a system or not.
Each is good in its own way. But no system can be everything to everybody.
===========
STV uses back-up preferences as a contingency vote.
But when used effectively to elect someone, that result varies from how first-choice votes were cast.
Under STV, the number of seats won per party is not necessarily proportional to first-choice votes per party.
But the most popular are elected, and few votes are wasted.
The same observation would apply under your party vote second-choice contingency system - as long as the voter chooses a party that is over threshold for his first or second vote. If not, no luck. the vote would still be wasted.
The contingency second vote would be used only for those who marked first choice for a party not over threshold.
If unlimited back-up preferences were allowed, the number of counts would be very small - with each successive choice used to place individual ballots, eventually the voter would mark a choice for a party over threshold and each count, the number still remaining in un-allocated pile would grow thinner and thinner.
Same as under STV, back-up preferences are only looked at if the first one on that individual ballot are unable to be used.
No big problem.
But note the result in the end after transfers according to second choices would not be proportional to the first choice votes.
But then the result under NZ's MMP is not proportional either (8 percent in last election did not get any representation) so it would be no worse that way than currently. (But it would be no better either, unless the second-choice votes on some ballots brought up a party from under threshold to over threshold - would you allow a party to get over threshold on transfers?)
But if one back-up preference is good for the party vote, why not for the district vote? If PR (or mixed proportionate rep or balanced rep or whatever you want to call it) is good at the national level, why not in each major cities through MM district(s) and transferable votes -- as under STV?
STV is used in NZ municipal elections so I know NZ knows of it.
Australia's STV where voters are forced to rank all candidates or mark party slates is not only form out there.
Pick one you like better if you don't like Australia's.
or stay with MMP, whatever.
PR of whatever form is better than FPTP, for sure.
========================
How do back-up preferences work? Do they lead to proportional results?
It is like a family is stuck in isolation and they ask a neighbour to pick up some baked goods and some dairy goods and some bananasa if they look good.
They tell him get some cream but if you can't get cream, get milk but not-skimmed milk.
get some bread, or buns anyway,
and some jelly donuts, but if you can't get jelly, no apple fritters, try for glazed donuts.
The neighbour comes back with un-skimmed milk, bread and glazed donuts, but couldn’t find bananas that looked good
The family is mostly happy -- they got some of what they most preferred but they did not get all of their first preferences, and they got nothing they had told the neighbour to not buy for sure.
They are happy - mostly anyway - but could we say the result was proportional? Hard to say.
Maybe we are using the wrong term - proportional representation.
Maybe PR does not quite get across the meaning we want, especially if we have a system that uses vote transfers such as a second-choice party vote system in MMP or STV.
But party-proportions is useful as a way to measure representation that reflects votes cast.
Another way is to look at how many voters helped to elect the members, what percentage of votes cast were actually used to elect someone. where each voter only casts one vote in a contest, that number is easy to assemble - whether to use all votes cast for most-popular candidate - or only those that give him or her a slight lead over nearest contender - is a different issue.
Catherine Helen Spence, Australia's leading lady of electoral reform 1890s to 1910, used to call for Effective Voting.
We all want our vote to be effective, and systems that let most votes be used are generally acclaimed as good systems.
So using the term Effective Voting makes good sense.
But the term “proportional representation” appears to be here to stay.
The term "proportional representation" is both useful as a shorthand and for representing the general idea of proportionality, which is equivalent to fairness.
But the term is is flawed for two reasons:
it has always been regarded as having the drawback of being polysyllabic
"proportional representation" are two big words put together
and now perhaps some people take the name too literally - those who think only first choice votes are all that is important to voters -
while others take it too loosely -- those who think it only means representation by population. (see footnote)
===========================================
Some say that Conservative party would have taken 138 seats our of 338 seats under PR.
But I don’t see this if the PR system is very proportional.
Taking the percentage of vote the Conservative received and multiplying that by the number of seats (338), you actually get the figure of 115 seats, a few less than the Conservatives got in the last election.
The fact the Liberal got more seats than the Conservatives in the 2021 eleciton is not due to under-representation of Conservatives but to over-representation of Liberals.
Any PR system is not strictly proportional, so sure it is possible that Conservatives might have taken 138 seats under PR. It should not take that many if the result was very proportional (based on votes as cast in 2021) but it could happen.
Note that 138 seats is not a majority, and that likely the party in power, under PR just as under FPTP, would be the Liberals, propped up by the NDP. Under FPTP and likely too under PR, the Libs/NDP combo would have a majority of the seats arising from a majority of the votes.
Proportional representation - proportional to what?
Only if every voter had a seat would representation be exactly proportional.
Any system that increases the seats or makes more votes count (effective voting or equal voting) is a step toward more proportionality.
Many European jurisdictions have fewer voters per member than Canadian systems.
City councils exceeding 50 members is not unknown in Europe, while anything larger than 20 is extremely rare in Canada
FPTP is not proportional -
It is doubly flawed -
only one member is elected in each district, therefore only one group can have representation in each district
and
as few as 18 percent of the vote in a district can elect that member.
Multiple-member districts or some other pooling of votes is required to have PR.
In my recent examination of mixed MMP and STV,
looking at the last federal election, we see no party taking a majority of votes
and under MMP/STV based on House of Commons with 386 seats, we see that the Conservatives with slightly more votes than the Liberals receive more seats – Conservatives 137 seats to Liberals’ 130 seats.
Under strict proportionally in 386-seat HofC, the result would be 131 Cons to 126 Liberal.
This is shown in the table below, in the underlined column
STV, even with MMP top-up, gives an extra little bit to the leading parties. so likely result would be Conservatives 137 seats to Liberals 130 seats. as shown in right column in bold.
But this result is more proportional (based on X voting, or first choice votes) than the FPTP 2021 election results, where the Liberals got more seats than the Conservatives although taking fewer votes.
Overall party seat tallies under MMP/STV By Province
Seats party
if strict vote
Maritimes QU ON MN SK AB BC Terr. TOTAL PR percent
Liberal 21 29 55 4 1 5 13 2 130 126 32.6
Cons. 12 17 49 7 11 25 16 0 137 131 33.7
BQ 0 35 0 0 0 0 0 0 35 30 7.6
NDP 4 9 25 4 2 7 14 1 66 70 17.8
Green 3 0 0 0 0 0 3 0 6 9 2.3
People's 7 0 0 1 2 2 0 0 12 20 5
37 90 139 15 15 39 48 3 386 386
Dist. seats 30 78 115 13 13 31 40 3 323
Top-up 7 12 24 2 2 8 8 63
We see the Maritimes and each other province electing a mixture of members from a variety of parties.
The table shows every province listed, except Ontario, electing members of four parties.
And Ontario may actually elect more than the three parties shown - if the Green party does better in Toronto than it did in 2021 (which is likely under PR) or if the Peoples’ Party does better in the rural area under PR, which also is likely, one of those two parties or both would likely take at least one seats thus giving Ontario the same diverse representation seen in other provinces.
Each province in the Maritimes is also likely to have more diversity of representation than they currently get under FPTP.
==============================================
MMP/STV Would End Artificial Regionalism
MMP/STV (or just STV for that matter) would have the virtue that each province (and even many cities) would elect a mixture of members, thus preventing the artificially created regionalism that now is so strong.
The result of the STV district election contests in the multi-member districts would not be necessarily proportional to first choices cast by voters but instead proportional to a mixture of effective first choice votes (those cast for successful candidates) combined with effective back-up preferences (votes initially cast for unsuccessful candidates and then transferred so that they could be used to help elect someone preferred by the voter over other candidates even if not the first choice of the voter).
The benefit of the vote transfers is to reduce the number of votes wasted and votes ignored - which under FPTP is often half of the votes cast across an election and even as many as 82 percent in individual district contests.
And that is good enough for me.
========
* Some say that rep by pop. is proportional representation - but not many think that way now. Maybe at one time, rep by pop. was a big step over unequal representation but now most of the world has moved to the next step of making each vote count through PR.
For example, online
we see this
"Is the US a proportional representation?
The Constitution provides for proportional representation in the U.S. House of Representatives and the seats in the House are apportioned based on state population according to the constitutionally mandated Census."
That’s not the way most people today think of the term proportional representation…
=========================================================
Comments