You may have been surprised at the outcome of the recent British elections, but New Zealand’s experience shows you should not have been surprised that you were surprised

While writing my history of New Zealand, I wondered about whether it would be possible to assess people’s attitudes before there were surveys. Writers often impose their prejudices, without realising they are doing so. That applies to contemporary writing of the times, while private writings – say, correspondence and diaries – reflect the musings of a particular (the most literate) part of the population (and often have little to say about the political questions my history is interested in).

I don’t trust newspapers; I have worked with too many journalists who are cynical about their readership and impose their prejudices on public opinion.

It occurred to me that every three years or so the government surveys the adult population with one question which should throw some light on the issues troubling me. I thought election voting might shed some light on the public’s attitudes about the political economy.

There were problems of interpretation. One is tactical voting, in which you might prefer Party C but voted for Party A because you wanted to keep out Party B. Then what to do about miscellaneous parties especially the ‘Non-vote’ Party which Bob Chapman pointed out often has an enormous impact on the elections outcomes (the less imaginative refer to it as ‘voter turnout’)? Additionally there was no proper conservative party until 1908 when Reform was created by Bill Massey so I could not really go back before then.

The elections, which occurred before there were useful surveys, were run under what is conventionally, but inaptly, called First Past the Post (FPP) regimes. There is no ‘post’, so the book refers to them as ‘Front Runner’ (FR); when the music stops, it is the candidate in front who is elected, even though he or she may have made little progress in the race.

What I found was that frequently the elected government of New Zealand did not reflect the wishes of the electorate. Historians tend to report about seats won, not about votes won. Sometimes there were significant divergences. Here are some examples.

1911 Election: The incumbent Liberals won 194,059 votes and Reform won 165,127. Yet Reform won 37 seats against the Liberals’ 33, becoming government after various deals and floor crossings. (Reform votes exceeded the Liberal vote in the next, 1914, election.)

1928 Election: This time the incumbent Reform Party won 263,382 votes but it was the ‘United/Liberals’) winning only 225,042 votes which became the government. Each won the same number of seats (27) but Ward acquired a few independents while Labour with 19 seats sat out the dance. Incidentally, between 1925 and 1928 Labour increased its voter support by 10,714 votes and won another 7 seats. In the previous triennium between 1922 and 1925 it had won 36,930 extra votes but lost 5 seats.

1935 Election: Seen as a great victory and a turning point in New Zealand politics, Labour won 46.1 percent of the vote and 53 seats, while the National Coalition won only 32.8 percent and only 19 seats. This was partly because two other parties (Country and Democrats) split from National, winning 10.3 percent of the vote, and fragmenting support. If you go through electorate by electorate adding the fragmented right-wing votes to National, the right would have won more seats than Labour. Once it was factionalism on the right which kept them out of power.

1946 Election: This was the last time a party won more than half the votes. Previous occasions were 1938 and the Liberals in 1890 to 1908 (excluding 1896), but then there was not really an organised alternative party and often elections ran under a system different from FR/FPP.

1951 Election: Here is what the draft book says:

            ‘[Following the ending of the Waterfront Dispute] the Labour Opposition had called for an election. Not a good move because it gave National the excuse to call a snap election, in which they gained another four seats. However their share of enrolled voters fell by 0.5 percentage points while Labour’s share fell 3.3 percentage points. The big ‘gain’ was those who did not vote, rising 3.4 percentage points. The outcome of the snap election is often portrayed as a ringing endorsement of National’s handling of the dispute; it is really an indication that the Labour movement handled matters worse.’

1978 and 1981 Elections: Both elections are noteworthy for Labour winning more votes but fewer seats than National.


1993 Election: Here is what the draft book says:

            ‘It was a very grumpy population that voted in 1993. National lost almost 200,000 votes compared to 1990, with its share of registered voters falling from 39.6 percent to 29.0 percent, lower than its support in the 1984 election after nine years of Muldoon. Yet Labour's share fell from 29.1 percent to 28.7 percent. The alternative left vote had consolidated into the Alliance whose main components were New Labour and the Greens; they won 15.1 percent of the vote, over a third of the left vote. Meanwhile New Zealand First, a breakaway centre-right party arising from Winston Peters leaving National, won 7.0 percent of the vote. ...‘It was the last FR/FPP general election; once more the allocation of parliamentary seats was eccentric. National won 50 seats, Labour 45, the Alliance 2 and New Zealand First 2.’

In about one in three elections under FR/FPP outcomes were markedly eccentric (space has meant leaving out some of the nuances) where parliamentary outcomes did not reflect voters’ wishes (even ignoring the Non-Vote Party). The effect was especially strong when there was a significant third party support reflecting that the populace’s opinions could no longer be treated as simply being on a left-right spectrum.

The surprise then is not that New Zealand switched to an alternative voting system but that it took so long to do so.

The May 2015 British election underscores the same lesson. Labour’s share of votes rose 1.5 percentage points, more than the Conservative’s 0.8 percent;, but it lost 26 seats while the Conservatives gained 24. At the specific level one can explain this by the complexity of the other third of voters switching between a multitude of minor parties. Conservatives got 36.9 percent of the vote, Labour 30.4 percent.*

As I watched the run-up to the election I wondered how anyone could predict the election outcome with confidence. I was not surprised the outcome was surprising.

Will Britain switch to a more representative voting system? It has a record of getting to the right solution only after it has tried everything else – often decades (or centuries) too late. The last opportunity for change was fumbled when the Liberal-Democrats got no movement on electoral reform despite being in coalition with the Conservatives as part of the previous government.

The problem arises because the FR/FPP electoral system evolved before there were parties as we understand them today. Electorates elected their members to represent them, not as representatives of some party. The result – as the New Zealand parliament before 1891 well illustrates – was that governments were rarely stable.

The creation of parties gave a stability – even an elected dictatorship. While there have been occasional bouts of turbulence, no New Zealand government has fallen because of instability since 1911. The price has been that MPs are usually loyal to their party rather than to their electorate.

Throughout the world the vast majority of today’s MPs give greater priority to their party. But our thinking and institutions have yet to adapt, particularly as our societies become increasingly heterogeneous. It took years after the establishment of the first party (the Liberals in 1891) for parliament to formally recognise the existence of parties. The first step I know of was legislation in the early 1970s when party allegiance was put on the ballot paper; the biggest changes occurred AFTER the referendum which adopted MMP. 

As for my general history? I think I was able to make some long run inferences from the electoral record – but weak ones. You’ll have to await its publication to find out.


* The Greens won one seat with 3.8 percent of the votes; SNP won 56 with 4.7 percent of the vote; UKIP with 12.6 percent got one seat, the same as the Greens; the Lib-Dems won 7.9 percent and got 8 seats. Each with 0.6 percent of the vote, the Democrat Unionists won 8 seats but Sinn Fein only 4. (The voter shares come from a BBC website. I have seen slightly different shares quoted elsewhere.)

Comments (6)

by Andrew P Nichols on May 13, 2015
Andrew P Nichols

I don’t trust newspapers; I have worked with too many journalists who are cynical about their readership and impose their prejudices on public opinion.

In my experience it's usually a self censorship reflecting not their own views but those of the editor appointed by one the dwindling number of far right oligarchs that own the media. They know that its not good for their future employment in a declining industry to be honest - unless of course they really do believe the crap they write.

by Moz on May 13, 2015

It's quite possibly chaotic in the mathematical sense - deterministic but extremely sensitive to input perturbation. I wonder if it would make a good example to use for baby maths students struggling with the usual examples.

Andrew: there's a strong selection effect, as you imply. Rather than overtly dictating what can be written the bought media hire and reward the tone and slant that they want. The same strategy has worked wonders in science, albeit with more effective protest from those afflicted.

by Rich on May 13, 2015

Andrew: how do you explain how the NZ Herald is right-wing whilst the (London) Independent is liberal/centrist, given that (until recently) they had the same owner? (Tony O'Reilly)

by Rich on May 13, 2015

(During the late 2000's, that was. O'Reilly has since sold both the Independent and most of APN).

by Sam Crawley on May 13, 2015
Sam Crawley

Some of the surprise about the UK result came not from the unpredictability of FPP, but from the size of the popular vote. It's true that tactical voting distorts that, but 6% difference between Labour and Conservatives was surprising given what all (well, most) of the polls had been showing.

That said, it's pretty obvious FPP has run its course in the UK. The electorate is more fragmented than ever, leading to some pretty disproportionate results (I would be interested to know how this election rates in terms of disproportionality compared to previous UK elections).

Despite that, I can't see proportional representation being adopted any time soon. For one, everyone is very distracted with the issues of Scotland and Europe (not to mention politics-as-usual). The UK, and particularly England, is conservative at heart (in the old sense of the world, rather than the "new-right" meaning). Labour and the Conservatives obviously won't want PR, and even if a referendum came about, the change movement would need to be driven by a grass-roots movement similar to the one in NZ, which will be a lot harder to build in the UK. Beyond that, the left (who traditionally support PR) may now be wondering if it's such a good idea, given the strength of UKIP.

I'm sure it will happen one day (perhaps as part of a deal to keep Scotland in the Union?), but probably not for at least another 20 years.

by Andrew P Nichols on May 14, 2015
Andrew P Nichols

Independent is liberal/centrist, given that (until recently) they had the same owner? (Tony O'Reilly) Thts;'s right It's owned by an expat  Russian oligarch now.

The Indy supported the Conservatives this election. It has a few worthy journos like Fisk to maintain a veneer of progressive independence but it still cheerfully promotes the western political agenda with respect to its ME wars and the Ukraine fiasco witbh all its tawdry propaganda.

Post new comment

You must be logged in to post a comment.