It is looking extremely hard for Labour to win in 2020. To do so it would need to do the impossible - move in two directions at the same time.
Taking a right turn, as represented by Liz Kendall, would win votes from Tory voters and probably win back a number of seats in the south of the country. However, it would effectively give up on any attempt to win back seats in Scotland, whose voters voted overwhelmingly for a left of centre, anti-austerity party. But this would not be the limit of the losses - Ukip came second in 120 seats, of which many were Labour seats in the north and midlands. A lot of these could go Ukip if Labour take its traditional voters for granted and tacks right. So a Kendall leadership would lead, at best, to the losses and gains balancing out, but could easily lead to a net reduction in seats.
Taking a left turn, as represented by Jeremy Corbyn, would result in the opposite scenario. Labour would likely take back a number of seats in Scotland and hold off the UKIP threat in the northern heartlands. But it would lose some Labour-Tory marginals, and would be unlikely to win too many seats back from the Conservatives. So once again, the result would be likely to be similar to this election; not the wipeout that the press imagine would take place under Corbyn, but not the transformative victory envisaged by the left either.
What of the two other candidates, Andy Burnham and Yvette Cooper? At the moment it is
difficult to ascertain what either of them stand for, but it seems that they are both Labour centrists, trying to tread a middle path that can hold the party together, allowing all wings of the party to feel included by being vague on a number of issues and relying on a large dose of ambiguity. In some ways they represent a continuation of Ed Miliband's approach, but with a more popular leader and less of the intellectual creativity. Under our current first past the post electoral system this is the most rational course of action; to win a party needs ideally 40% of the vote (David Cameron's 37%, like Tony Blair's 36% in 2005, is unusually low for a winning party). To reach that level of support a party must be a broad coalition of different traditions and views, none of which feels too unhappy with the inevitable policy compromises they have had to make on the way. This worked well in 1997 and 2001, largely because the Tories were tearing themselves apart after 18 years of power and because left-wing voters had nowhere else to go. But with the current strength of the SNP and of UKIP, such an approach no longer works. Even if we assume that in this election. Labour suffered due to an unpopular leader and would do better under either Burnham or Cooper, it's hard to see the 'Labour centrist' strategy getting the party much above 35%.
The problem is that it is very difficult for Labour to gain more support from one group of voters without losing support from another. Move right to coax in Tory voters and you lose more people to SNP, the Greens and to apathy. Try to sound tougher on immigration to attract UKIP voters and you lose liberal voters who are the bedrock of Labour's support in the big cities. Offer a stronger anti-austerity platform, and you risk lose centrist voters to the Liberal Democrats and to the Tories. Trying to satisfy everyone at the same time proves the most dangerous path of all - you lose any sense of authenticity or vision, and are forced to rely on platitudes, truisms and a constant refusal to articulate what your policies would be.
If neither moving left, moving right or staying where they are offer Labour a path back to government what strategy should they adopt? The only way forward is to embrace proportional representation. PR has much to recommend in terms of fairness but it offers something very specific to the Labour party; the opportunity to split into two parties without letting the tories in. Under our current system, a party is unlikely to win many seats unless it has 25+% of the vote (unless its support is highly concentrated in particular areas). So under First Past the Post a split is fatal, as Labour and the SDP learned in 1983. Under proportional representation however, seats are allocated roughly in proportion with the votes cast. Which means that Labour could split into a Jeremy Corbyn-style socialist party, attracting up to 20% of the vote, and a Kendall-type centrist party, attracting 20-25%. The two parties would compete at elections, each being able to speak authentically and with passion, free from the straitjacket of trying to reach out to totally divergent interest groups. As a result the total vote share gained by the two parties would almost certainly be greater than they could have achieved as a unified party. After an election they would be able to form a coalition together, deciding their policy platform based upon their two manifestos and on which had proved more popular with the electorate.
Labour has generally favoured retaining first past the post; for the simple reason that the party thought the system could grant them majority governments. That is no longer the case. It was always difficult for Labour to win under first past the post - in the current state of multi party politics it has become almost impossible. It's time for the party to face up to electoral reality - the leadership candidates need to put proportional representation onto the agenda.