One of the intriguing ‘what if?’ questions of recent British politics is what might have been had Tony Blair not left Paddy Ashdown at the altar but brought Labour and the Liberal Democrats together in government in 1997.

‘It would have been a better, more progressive government,’ says ex-Downing Street adviser Roger Liddle, among those to leave Labour for the SDP before returning to New Labour. Constitutional change would have been deeper and more coherent. On Europe, the government would have been ‘much less in hock to Rupert Murdoch’. That could even have changed the multilateral diplomacy over Iraq, too, Liddle suggests.

This could be wishful thinking: a Lab-Lib government could have broken down over Iraq or another disagreement. The detailed debates of parallel universes are never fully available to us. But the political reforms which have returned to consideration late in Labour’s third term – an elected second chamber, electoral reform – would surely have been foundational to such a coalition.

The centre of gravity of British politics could have been redefined more confidently. Labour routed William Hague’s Conservatives in 2001, yet much of the loss of momentum of New Labour in power arose from an inability to ‘ratify’ victory in arguments which were central to that campaign – over Europe, public spending and taxation. The Tory inquest into successive election defeats might have begun earlier than 2005, and perhaps also gone rather deeper than the ambiguities of Cameronism. In large part, these mirror the lack of clear public definition of New Labour’s own legacy.

It didn’t happen. Party tribalism has a powerful gravitational pull. Perhaps Labour’s 1997 majority seemed too large for Blair to proceed, though a coalition of choice, not necessity, would have sent a powerful signal of pluralist intent. Instead, cooperation over constitutional affairs broke down following Labour’s shelving of the promised referendum after the Jenkins report on electoral reform.

Now the question of Labour-LibDem cooperation is stirring again in very different political conditions. It is discussed in thinktanks and newspaper columns, and sotto voce in the Commons bars and tearooms. But it seems far too late – and far too much a matter of low political survival rather than political principle.

Yet political cooperation across parties remains in the strategic interests of both parties and, more fundamentally, of the major progressive causes and ambitions for British society.

First, there is no fundamental ideological difference between the two parties of the centre-left, despite differences in history, ethos and culture, as well as more tactical questions of policy, positioning and electoral demographics for competing parties. Senior Labour and LibDem figures have engaged over the possibilities and barriers to progressive cooperation at Fabian Society and CentreForum events at both party conferences in recent years.

Second, policy differences are greater than they were a decade ago. But where the parties have different instincts, these are often necessary debates within each party which might benefit from mutual interrogation. If LibDems have cause to challenge Labour illiberalism, a counter-claim can be made about the dangers of too little emphasis on breaking down class-based disadvantage, and challenging inequalities of assets and wealth. (The LibDems would abolish the child trust fund). And progressive activists across all parties and none need to think harder about how to win public arguments on hard issues such as crime and immigration.

Third, there are significant political advantages to cooperation. The idea of a ‘progressive majority’ against Margaret Thatcher in the 1980s is often excessively simplistic. But the outcomes of British politics when progressives cooperate are likely to be more progressive.

Still, there is little or no chance now of restoring starry-eyed Lab-Lib cooperation at this stage of the political cycle. The paradox of political cooperation is this: it will succeed when done from a position of political strength, yet parties are often tempted to act only when too weak to carry it off.

It is interesting how far the Labour politics have shifted. This was a Blairite project, which raised John Prescott’s hackles. Jon Cruddas, who topped the deputy leadership poll in the first round, recently wrote: ‘Constitutional and electoral reform requires an alliance with the Liberal Democrats so let’s make one – socialists and social liberals hold much in common.’ The politics of cooperation no longer fits the same left-right tramlines. Progress’ campaign for a citizens’ convention has found growing support within New Labour while Compass’ campaign for a broad left front is a similar cause.

Most of the barriers are now on the LibDem side: once bitten, twice shy. And they have perhaps been nibbled rather more often than that – including in Gordon Brown’s ‘government of all the talents’ gambit in 2007. (The offer of individual LibDem cabinet members without any party arrangement demonstrated the trade-off between the politics of pluralism and the politics of control). So there is little appetite to repeat the experience of the Lab-Lib pact of the 1970s. If cross-party anti-Tory voting was a considerable feature of the 1997 election – a kind of ‘do-it-yourself’ Alternative Vote – and still a factor in 2001 and 2005, the elections of 2009 showed an ‘anybody but Labour’ factor at work. Many LibDems regard Labour as a contaminated brand in British politics: close engagement with a 12-year-old government as undermining a message of change and reform.

Yet the LibDem edge over anti-political parties depends on a credible claim that they could make reform happen. That depends on a politics of cooperation – and remains much more likely on left than right. (Look at the visceral Tory reaction to electoral reform being entertained). Indeed, its strategic goal is to make pluralism the norm in British politics. That can’t be achieved by purist insurgency.

In the long run, we will need to cooperate. Yet, as John Maynard Keynes noted, in the long run, we are all dead. He was a liberal whose responses to the Great Depression were rejected by Labourism and Treasury convention.

So what might be done even now? First, get on the front foot on major progressive causes. Political reform is central. The Cameron agenda is cosmetic. There is a good deal of overlap between Progress’ campaign and Nick Clegg’s political reform package. Giving away the power to decide to a citizens’ assembly is the best way to deal with the charge of a last ditch change of mind.

Second, go into a difficult election with a manifesto that does not contain ‘red lines’ against cross-party cooperation should a hung parliament result. Better to change policies before meeting the voters than afterwards. Goodbye to ID cards, with little lamentation. If these policy changes cheer the party’s constructive left, surely Blairites would think it a price worth paying for their ‘project’.

Third, and perhaps most controversially, without any question of electoral pacts, should unilateral action be considered? Is it in Labour’s interest to contest those 30 or so LibDem seats being defended against strong Conservative challenges where Labour is not in contention? If Labour recovers strongly, these seats could be crucial in determining whether the Tories can gain a majority, or who has most seats if there is a hung parliament. There is a good argument against retreating from territory: however, I think a case can be made for doing so exceptionally in one last first-past-the-post contest if combined with advocacy of a new electoral system which would do away with tactical voting.

The politics of progressive cooperation will certainly return – by the London elections of 2012, if not before. It is never too late to start.