Northern Ireland at 100: Unionism failing; Nationalism stuck; Moderates thriving…

One hundred years after Partition, Northern Ireland is still in existence. It would surely come as a big surprise to many who thronged the streets of Belfast on June 22nd 1921 – the date King George V opened the first NI Parliament in City Hall – that unionism is now a minority in Stormont. It would also surely come as a big surprise to many nationalists in 1972 that the state set up to guarantee unionist rule in north-east Ireland is still in existence. The Queen speaks Irish in Dublin Castle; Martin McGuinness toasts the Queen in Windsor Castle. Has everything changed? Has anything changed?

The demography of Northern Ireland has changed, certainly. In the 1918 Westminster and 1921 NI elections, unionist candidates gained 65.9% and 66.9%, respectively. Nationalist figures were 30.8% and 32.4%; and middle-ground figures were 3.3% and 0.8%. Abolition of PR-STV for NI elections in the 1920s meant that there were many uncontested seats in both Stormont and Westminster until 1970.

The three graphs below were created using a 5-election average, with the date being the average date of the five elections. Because there were so few elections when every seat was contested, the 5-election average figures pre-1971 should be treated with more caution than more recent averages. The most recent set of points on the graphs represent a 3-election average, and the second most recent set of points represent a 4-election average: this is to allow a smoothing to be applied to very recent elections.

Candidates were classified based on transfer analysis. The vast majority of Westminster candidates have also stood in the other elections so they are classified without much difficulty.

Each election type has its own unique idiosyncrasies. Pro-European nationalist voters typically had a much larger turnout than unionists; and Westminster FPTP elections give rise to electoral pacts and tactical voting. A 5-election averaging can smooth out these psephological ripples and let us see the real patterns emerging.


5-election smoothing of Unionist bloc vote in Northern Ireland: 1921-2021.

What is unionism’s real pattern? Six stages can be discerned.

  1. Unchallenged Hegemony: 1921 to the mid-1960s, when the challenge came from NILP (about 2% drop over 44 years).
  2. Slight Wobble: 1965-1968, when nationalist and NILP reinvigoration at the polls reduced unionist share to the low 60s (4% drop over 3 years).
  3. High Turnout but Fragmented: 1968-1975, unionist politics splinters but turnout stays high (1% drop over 5 years).
  4. Struggling against Demography: 1975 to 1998, a slow but inexorable decline in unionist vote share to about 50% (9% drop over 23 years).
  5. Post-GFA Dividend: 1998 to 2016, differential abstention rate increases among nationalists giving a relative flattening of the unionist vote share. Yet the unionist vote dips below 50% (2% drop over 18 years).
  6. Feeding the Crocodile, RHI and Brexit: 2016 to 2019, a rapid drop in unionist vote share: 5.5% drop in under 4 years. Nationalist voters flock to the polls, and some unionists jump ship to Alliance.

Religious demographic change is making life difficult for unionists, but demography doesn’t cause either a short-term flattening or a steep drop in unionist vote share. The conclusion, based on post-GFA vote share, is that making Northern Ireland work efficiently in a way that doesn’t antagonize any section of the community is the best way to maximise unionist vote share.


5-election smoothing of Nationalist bloc vote in Northern Ireland: 1921-2021.

Three phases in nationalism’s vote pattern can be seen since the early 1990s:

  1. Surfing the Demographic Tide: 1991 to 1998, a rise from the mid-30s to 40%.
  2. Plateau: 1998 to 2003, the nationalist vote share plateaus at its highest ever (42.4%).
  3. Stagnation and Decline: 2003 to 2019, slow decline in the nationalist vote share to 40.5%.

On the basis of these voting figures, the likely 2021 census plurality of Catholics is not a basis for calling for a referendum on a United Ireland, as the overall nationalist vote (SF, SDLP, WP, PBP, Aontú) has been declining since 2003.


5-election smoothing of Middle-Ground bloc vote in Northern Ireland: 1921-2021.

And what of the middle-ground? Vote share seems to take four phases:

  1. NILP Growth: up to the late 1960s.
  2. Alliance Arrives but no Takeoff: 1970 to 2003, polarization makes Alliance breakthrough difficult (Alliance did very well in local elections in the early 70s, but less well in Westminster or Assembly elections).
  3. Slow Progress: 2003 to 2016.
  4. A Plague on Both Your Houses: 2016 to 2019, both blocs, but particularly unionism, lose votes to Alliance and the Greens.

Middle-ground politics has reached its current height before, in the mid-60s. What is different now? Many voters, particularly those born after 1998, define themselves as Northern Irish rather than exclusively Irish or British.


What these three graphs tell us is that unionism is in the biggest trouble, having to contend with demographics and a reinvigorated middle-ground. Nationalism seems unable to capitalize on the growth of the Catholic share of the population, never mind attract support from a sizeable section of Protestant or Other/No Religion voters. Middle-ground politics is thriving among those who refuse simplistic identity labels.

The exclusionary identities of nationalism and unionism are failing to command consensual allegiance in NI. Nationalism and unionism have often relied on each other to re-energise their base, but this strategy is not working now when the centre-ground is so large. Indeed, such a strategy seems to be helping Alliance and the Greens. This may be a post-GFA effect: if ‘moderate partisans’ accept the power-sharing political paradigm that is the GFA, they may be less likely to scuttle back to their originatory sectarian redoubts when attempts to inflame politics occur.

Both exclusionary identities will have to widen to become inclusionary identities in order to grow. Both identities need to embrace the other, not just in a desultory fashion, but as inherent to their own identity.

Such a task will be extremely difficult for unionism to achieve. The problem with loyalty to a Protestant Monarch and a belief in the integrity of the United Kingdom is that these are not things within NI unionism’s control. What if Prince Charles becomes a Catholic? What if Scotland or Wales become independent? Unionism needs to develop an Ulster-Britishness that is coherent and broad-church enough to absorb the potential seismic threats to its identity from Great Britain – and which is attractive to a significant number of NI Catholics – if it wants NI to survive as a political entity separate from the Irish Republic. At the same time, unionists need to – along with Welsh and Scottish unionists – reimagine Britishness so that it can withstand English threats to the Union. In di Lampedusa’s The Leopard, Tancredi could have been describing unionism: “If we want things to stay as they are, things will have to change”.

How could nationalism embrace British identity? Looking at vote transfers in PR-STV elections, the vast majority of unionists and middle-ground voters are not attracted to post-militant Northern separatism. Many nationalists seem to believe the opposite of Tancredi: that, by not changing any core belief or symbolism of their identity, things will change in the North simply through demography. The recent LucidTalk poll for BBC Spotlight shows quite a bit of resistance to changing the flag or national anthem (question 8: NI and RoI). For nationalism to become an inclusionary identity, Dublin will have an important role. The South needs to embrace the British identity of many of its own inhabitants, for example Donegal or Monaghan Protestants. Rejoining the Commonwealth, removing the Angelus from public service broadcasting, and deleting the Preamble to the 1937 Constitution: these reforms would probably be viewed positively by NI middle-ground voters.

And what of middle-ground politics? Is it forever to be handmaid to the other two big behemoths, or is it becoming a coherent political identity in its own right? The latter seems to be the case, as the number who now identify as Northern Irish continues to grow. 

One hundred years ago, King George V, in “a great and critical occasion in the history of the Six Counties” appealed:

…to all Irishmen to pause, to stretch out the hand of forbearance and conciliation, to forgive and to forget, and to join in making for the land which they love a new era of peace, contentment, and goodwill…

Unionism and nationalism, if they are to halt the slide in their electoral fortunes, could do worse than take such sentiments to heart. If NI survives as part of the UK, it will be because middle-ground voters vote for it. If Irish unity prevails, that will have been made possible by the middle-ground electorate. Whichever identity displays the greatest devotion to peace, contentment and goodwill is likely to prevail.

Photo by GLopezR is licensed under CC BY-NC-SA