It’s difficult to describe the atmosphere outside Ireland’s parliament on the evening of 14 November 2012, less than 24 hours after the country had learned of the death of Savita Halappanavar.
We had all known that Irish abortion law was archaic and dangerous. But we’d never before faced such a clear instance of a woman dying in an Irish hospital, specifically because of the apathy and indifference of our government. Savita, suffering from complications, and told she would miscarry, asked for a termination but was refused. She died shortly after giving birth to a stillborn baby.
The demonstration that night was eerily quiet. A few drivers honked their horns, suddenly trapped in a near-spontaneous crowd of thousands, but no one responded in anger, no one shouted at them. People just stayed there, in the road, with their candles, their pictures of Savita. Occasionally someone would try to start a chant, but it would soon get sucked into the silence.
We weren’t angry that night. We were numb, grief-stricken and deeply, achingly ashamed.
But very soon afterwards, Irish women and allies found their anger. And through five years of relentless campaigning – of demanding to be heard and rejecting the half-measures of government – we have won a referendum on the “cruel, inhuman and degrading” Eighth Amendment to our constitution.
Unfortunately, that was the easy part.
Campaigners, who up until now have simply been battling to make the referendum happen, now have to figure out how to win it. And the odds don’t seem to be in our favour.
Polling shows that a clear majority support a softening of the law in situations of rape (76 per cent), fatal foetal abnormality (67 per cent) or danger to the physical (82 per cent) or mental (72 per cent) health of the mother. But a large majority (67 per cent) still oppose abortion “on request”.
These numbers present us with tough choices. The simplest way to win the referendum would be to run a campaign entirely based on extreme cases of rape, incest, severe illness or fatal abnormality. We know that the Irish public is responsive to the harrowing stories of the most vulnerable women.
But the most vocal feminist activists are, unsurprisingly, pro-choice in a much wider range of situations. They’re in the 33 per cent who support abortion on request and many – quite reasonably – believe that to defend abortion only in the most extreme cases is to undermine the inviolability of every woman’s choice.
So where do we draw the line? To what extent can we compromise our feminism for the right result? Should we speak frankly about the fact that Ireland’s abortion law is deeply classist and racist, since only European-born and relatively well-off women can easily travel abroad for terminations?
In 2015, the pro-marriage equality campaign faced the same questions and opted for the most conservative approach possible. And they had a much clearer poll lead than the pro-choice campaign does, and a much easier sell.
Because no matter how much Repeal campaigners believe in the justice of the cause, we have to accept that if we win this referendum, the margin could be vanishingly small. In 1995, divorce was legalised with less than 10,000 votes to spare.
So, over the next 10 months, campaigners need to gather the best data they possibly can, identifying the key segments of swing voters. And it may turn out that to win those voters, we need to offer a message that undermines what many of us believe. We may have to swallow our anger and present a docile, non-threatening face to the country.
Of course, this conundrum is nothing new. It’s one that all progressives must grapple with every time they go to the polls.
But in this campaign, Irish feminists are carrying that memory of November 2012. They’re carrying the knowledge that if this referendum goes the wrong way, another generation of women – many of whom have yet to be born – will face the same terrible risks and indignities as Savita Halappanavar and countless others.
So in the months ahead, we have huge questions to answer. And we can’t afford to get them wrong.