Religious-political mapping in Westminster elections

Regular reader of this blog will know that one of its basic premises is that there is a close correlation between religion (or 'community') and political preference in Northern Ireland.

To be simplistic, this blog considers that the vast majority of (cultural) Catholics vote for nationalist parties, and that the vast majority of (cultural) Protestants vote for unionist parties. (As a consequence, of course, this blog believes that the relative growth of the Catholic community in Northern Ireland will lead to a consequent growth in the nationalist share of the vote).

But is the basic premise correct?

One way to test the premise is to compare the outcome of the most recent (2001) census with the outcomes of recent elections. As an election for the British parliament is imminent, it is interesting to look at the mapping of community identification with political identification in each of Northern Ireland's 18 Westminster constituencies.

There was a Westminster election in 2001 – the same year as the census – so statistics on the community identification of the electorate in each constituency at the exact time of the election are available. The table below gives, for each Westminster constituency, the religious identification of the electorate in 2001 (Census Table S306: Age by Community Background (Religion or Religion brought up in)), and the outcome of the election:

Several things stand out from the table:

• In most constituencies there was a close correlation between the proportions of cultural Catholics and nationalists, and between the proportions of cultural Protestants and unionists. Where there was no strong 'interference' from the Alliance Party or significant 'others' (e.g. in South Belfast – the Women's Coalition) the correlation can be very close indeed: East Derry, Fermanagh and South Tyrone, Newry and Armagh, North Antrim and South Down are examples of this.
• Where the Alliance Party won a significant score, the nationalist score was lower than the Catholic community proportion, indicating that the APNI picked up some Catholic community votes: East Antrim, South Antrim, Lagan Valley and Strangford are examples of this. However, as Lagan Valley and East Belfast show, the APNI also picked up Protestant community votes in similar numbers.
• There is clear evidence of strategic voting by cultural Protestants in several constituencies, generally to try to stave off a Sinn Féin victory: West Belfast, Foyle, Mid Ulster, Newry and Armagh and West Tyrone are examples. On the other hand, it seems as if some cultural Catholics may have voted strategically in South Antrim in a successful attempt to avoid a DUP win.

What can such an analysis tell us about the likely outcome of this year's Westminster elections?

• Firstly, although voters seem to vote largely in predictable political-religious patterns, they are more ready to vote strategically than often thought. Despite the absence of anything other than a 'first preference' option in First-past-the-post elections, some voters are prepared to anticipate the failure of their first preference, and vote for a candidate most likely to beat their least-favourite. This tendency is visible on both sides of the community divide, and may affect the outcome in several constituencies in 2010; particularly South Down, Foyle, Fermanagh and South Tyrone and South Belfast. The failure (so far) of the unionist parties to agree on 'unionist unity' candidates may not stop individual unionists from deciding who their preferred 'unity' candidate is.
• It is hard to judge whether the DUP or the TUV would be nationalists' 'least-favourite' – in North Antrim this could prove to be important, especially if the two right-wing unionist parties are neck-and-neck in the constituency in opinion polls. UCUNF in this constituency have effectively withdrawn, by standing an unknown newcomer, so some of their votes many also be in play.
• Where the experiences of 2001 and 2005 have showed unionists that they have no chance of ousting a strong Sinn Féin incumbent, either directly or by strategically voting for an SDLP candidate, they may revert to simply 'flying the flag' – voting for a dead-cert unionist loser just to show their colours. This may occur in West Tyrone, Mid Ulster, Newry and Armagh and West Belfast. The votes received by the SDLP in these areas may suffer as a result.
• The Catholic community voters in South Antrim, who were prepared to vote for the UUP's David Burnside in 2001 (and to a lesser extent in 2005) in order to try to thwart the despised William McCrea of the DUP, may be persuadable again. This may be one of the reasons why the UCUNF is being very careful about its choice of candidate – the right person could win them the seat with the help of strategic Catholics, but a known bigot like Adrian Watson would not.
• In 2001, and more so in 2005, David Trimble probably also benefitted from cultural Catholic strategic votes in Upper Bann – but whether his descendent Harry Hamilton can do so is less likely. If these votes revert to nationalist candidates, then this seat becomes winnable for Sinn Féin.

As in all elections, pre-match analysis tends to be backwards looking, and the actual outcome will not be exactly as predicted. The choices made by individual voters on the day will differ from the predictions in many respects, and will depend on who is standing and which party seems to have a chance. The strategic voting that will take place is even more complicated this time as there are, in some constituencies, three unionist parties. The pattern of voting in 2001 does not disprove the premise of this blog, but it does demonstrate that voters make compromises in the privacy of the polling booth, and as a result no outcomes can be excluded out-of-hand.