A lot has happened since June. For one, the NZ Labour Party has a charismatic new leader. The rise of Jacinda Ardern is causing a lot of excitement in the media. But there is good reason to be skeptical of what Bryce Edwards calls ‘Jacindarama’.

I would argue that the main challenge for Labour is to win back the working-class. Last year, I did some research into this for my MA thesis. I have summarised my findings on the University of Auckland Politics & IR blog Pacific Outlier. The following table illustrates my argument.

	Working-class		Middle-class
	Labour	National	Labour	National
1996	32%	29%	26%	37%
1999	47%	22%	32%	37%
2002	45%	16%	40%	23%
2005	47%	32%	42%	38%
2008	42%	38%	28%	50%
2011	39%	36%	20%	54%
2014	35%	38%	20%	52%

These numbers come from the NZ Election Study. I explain my methodology in the Pacific Outlier post, but basically the working-class are those voters in manual and low or semi-skilled non-manual employment. It includes both the traditional blue-collar workforce and those in routine white-collar employment. I estimated that, in 2014, the working-class made up nearly 40 percent of the vote.

But it was National, the party of businessmen and farmers, that won the working-class in 2014. A further 18 percent went to the Greens and NZ First (nine percent each). This is significant because until now, Labour has relied on strong working-class support to carry it to victory. The Labour share of the working-class vote has fallen from an average of 46 percent in 1999-2005 to 35 percent in 2014.

To put it another way, Labour must win back the working-class before it can win the country.

The Listener has published the results of a poll conducted 19-24^th May. In contrast to other major polls, this one has included undecided voters in the base of the calculation. That means the percentage of party support is considerably lower than it would otherwise be. Whereas 44% of decided voters support National, only 35% of the overall sample do (see this table shared by Matthew Hooton). It is the latter figure that The Listener has chosen to publish. But as we know, a large proportion of those eligible to vote choose not to.

In the last election, turnout was only 77%. That is why pollsters leave out the undecideds when calculating party support. After all, we don’t calculate the party vote as a percentage of the electorate. If we did, then the National vote was only 36% in 2014. That makes The Listener poll less sensational than the headline numbers suggest. And it means the result is consistent with every other poll we have seen out this year. National leads Labour by a wide margin and any centre-left government would depend on the support of NZ First.

But there is one major difference: the new poll suggests that Labour’s electoral support has not recovered from 2014. Only 19% of respondents would vote Labour. If we remove the undecideds, this would mean a party vote of 24% – one percentage point less than the 2014 result.

And yet, until now, Labour has been averaging 29-30%. Last week, these poll numbers were cited by NZ Herald journalist Claire Trevett as evidence that Labour would avoid further collapse. Trevett went on to suggest that any failure of Labour to form a government this year could have more to do with the unreliability of Winston Peters than the Labour vote. This is a naïve claim.

In May 2014, Labour was polling slightly better than it is now.  But after four months of campaigning, it finished with barely a quarter of the vote. Analysis by David Farrar on Kiwiblog suggests that this has been a recurring theme in every general election since 1999. On average, Labour support has declined 5% from the start of election year to election day.

It may be that the worst is yet to come for Labour. Certainly, political columnist John Armstrong agrees. The haphazard and lackluster response of Andrew Little to the Budget has prompted Armstrong to argue that the greatest challenge for Labour in 2017 could be preserving its status as a major political party. The Listener poll gives further justification to this argument.

Until a fortnight ago, the British Labour Party was on course to a major defeat. Early forecasts suggested it would lose up to 75 MPs. That would represent the party’s worst electoral result since 1935. And it would confirm a long-term decline: Labour has lost seats at every general election since 1997.

In 2015, Labour’s share of the national vote increased marginally from 29% to 30.5%. But it still lost a further 26 constituencies to the Conservatives and the Scottish National Party. Two years’ later, the Conservative Party was far ahead in the polls and certain to increase its majority, while Labour appeared to have imploded under the leadership of Jeremy Corbyn.

Not anymore.

With only days to go, opinion polls have swung dramatically against the Conservatives.  The Economist has aggregated all the major polls into an interactive graph. As of 1 June, the Conservatives were averaging 43% and Labour 37%. That compares to 49% and 26% at the beginning of the campaign. In other words, Labour has reduced the Conservatives’ lead by nearly two-thirds.

These numbers don’t give us the whole picture though. If we look at who is supporting the Conservatives, we find a worrying trend for Labour. The Economist has very crudely made a distinction between ‘rich’ and ‘poor’ based on occupation. The latter are those in skilled and unskilled manual employment. These voters represent Labour’s traditional working-class base and make up a formidable voting bloc.

Yet, we find that more of them support the Conservatives than do Labour. This is a new phenomenon. In 2015, based on an Ipsos MORI poll, Labour won 36% of manual occupation voters and the Conservatives 30%. As of 1 June, the Conservatives were averaging 43% (+13) among this demographic, while Labour was on 38% (+1). For the first time in history, the Conservatives are going to win the working-class vote, helped in large part by the collapse of UKIP.

That is the main reason I think Corbyn won’t be PM, and why the British Labour Party remains in serious trouble, despite a surge in the polls. Labour may increase its share of the vote, as it did in 2015, but it will fail to win enough seats in the working-class heartlands of North England and Scotland for it to have any chance of forming a government.

In a subsequent post, I will discuss what the failure of Corbynism means for NZ Labour.

The latest Roy Morgan poll records no substantial change in party support during the period 1-14 May. National remained on 43%, while Labour was down one percentage point to 28.5%.  According to Gary Morgan, the Executive Chairman of Roy Morgan Research, NZ First is still in a “strong position” to decide the next government.

But that was before the Budget. On Thursday, Finance Minister Steven Joyce announced a $2 billion “Family Incomes Package”. The government will increase financial support for an estimated 1.3 million families at an average rate of $26 per week. The adjustment of income tax thresholds also means that wage and salary earners, in general, will get some reprieve.

Morgan argues the Budget announcement could prove to be a ‘game-changer’ for National. And with 43% support, National would need only a modest shift in the polls to recover its electoral advantage over Labour and the Greens.

It is worth remembering that 43% is a mere percentage point less than what National got in the 2008 election, and only four percentage points less than the 2014 result. This means National hasn’t got a lot of ground to make up between now and September.

So, let’s suppose the Family Incomes Package is a game-changer.

Going on the Roy Morgan poll, 45% could be enough for National to govern without Winston Peters. Should Peter Dunne be re-elected in Ohariu, and should Marama Fox of the Maori Party be returned, the government would remain virtually unchanged.