Irish Times/Isbos-MRBI Opinion Poll June 11 2010: A Geographical Perspective

The Irish Times/Ispos MRBI poll, published in The Irish Times on Friday June 11th, rates Labour as the most popular political party in the state at present and also suggests that Sinn Féin will win that party’s highest share of the national vote since the 1920s. The poll figures offer a more sobering picture for Fine Gael, whose support levels now stand at just 1% higher than what the party won in 2007, while the threat of Electoral Armageddon hangs over the heads of the government parties, with Fianna Fáil likely to lose half of the its seats and the Green Party expected to lose all of the party’s six seats should these poll figures be replicated in a subsequent general election.

The proportion of seats won by parties in Irish general elections does not tend to measure up exactly to their actual share of the first preference vote (although the system is decidedly more proportional than its UK counterpart), mainly because party’s first preference votes need to be filtered through the system of Irish electoral constituencies. In order to address this question, I attempt to estimate what party first preference votes would be in different constituencies, assuming similar (proportional) change in party vote shares in all constituencies, as I did with the earlier RedC poll analyses. The only difference in this case is that my estimates are based on changing party supports levels regionally (based on party support by province figures referred to in the analysis by Stephen Collins) rather than nationally, given that regional breakdowns in support are, by and large, available for this poll. How does this work? Well, for instance, Fianna Fail’s share of the vote in Dublin is estimated in this poll to now stand at just two-sevenths of the level of support that the party won in 2007, while Labour support in the Dublin region is estimate to be over twice the vote share that the party won in the capital in the last election.  This of course is a very rough model, and ignores the fact that changing support levels between elections tend to vary geographically, even within specific provinces or regions, and also the impact that territory transfers, brought in by the 2007 Constituency Commission report, would have on vote share (although changing numbers of seats are taken account of). On the assumption that the base level of Labour support in Roscommon-South Leitrim and Mayo will have been added to by the recent acquisitions of John Kelly and Jerry Cowley (with the base level of Fianna Fáil support being similarly improved by the return of Beverley Flynn to the fold), I have added these candidates 2007 levels to the Labour (and Fianna Fáil) support levels in these constituencies to generate more realistic estimates there. Based on these estimated figures, I proceed to estimate the destination of seats in the different constituencies in each of the different region/provinces and then to aggregate these to complete the national estimates.

My estimates as to what parties’ shares of the first preference votes in the different constituencies would be:

  FF FG LB GP SF OTH
Carlow-Kilkenny 29.8% 33.1% 25.4% 6.0% 5.7% 0.0%
Cavan-Monaghan 24.3% 31.7% 4.7% 8.0% 23.0% 8.4%
Clare 24.6% 41.7% 5.1% 4.6% 8.2% 15.9%
Cork East 14.7% 25.2% 46.4% 1.8% 11.3% 0.6%
Cork North Central 14.6% 23.9% 29.0% 2.3% 14.3% 15.9%
Cork North West 31.5% 48.3% 16.7% 3.5% 0.0% 0.0%
Cork South Central 22.3% 30.3% 26.7% 6.8% 11.1% 2.9%
Cork South West 20.6% 36.9% 26.8% 5.2% 10.5% 0.0%
Donegal North East 34.3% 24.3% 7.5% 3.1% 21.3% 9.4%
Donegal South West 34.0% 24.4% 11.3% 3.4% 25.5% 1.4%
Dublin Central 16.3% 10.5% 33.6% 2.7% 15.3% 21.6%
Dublin Mid West 13.6% 14.9% 32.9% 5.7% 17.3% 15.6%
Dublin North 18.4% 18.5% 30.7% 9.3% 5.3% 17.8%
Dublin North Central 18.0% 31.5% 21.7% 2.7% 7.0% 19.0%
Dublin North East 13.8% 24.0% 38.3% 3.0% 21.0% 0.0%
Dublin North West 15.8% 9.7% 47.8% 1.1% 23.0% 2.5%
Dublin South 18.1% 36.0% 33.3% 6.2% 6.0% 0.4%
Dublin South Central 10.1% 13.2% 46.9% 2.3% 14.0% 13.5%
Dublin South East 11.1% 21.7% 46.8% 6.8% 8.3% 5.4%
Dublin South West 12.2% 18.8% 45.4% 1.5% 17.2% 4.9%
Dublin West 12.5% 20.6% 41.7% 1.6% 7.3% 16.4%
Dun Laoghaire 12.8% 26.2% 43.0% 3.6% 3.7% 10.8%
Galway East 26.9% 41.9% 12.9% 4.4% 3.9% 10.0%
Galway West 21.3% 18.4% 38.2% 10.7% 3.0% 8.3%
Kerry North-West Limerick 11.9% 26.0% 23.8% 1.2% 33.4% 3.7%
Kerry South 17.2% 22.5% 32.7% 1.3% 6.4% 19.9%
Kildare North 20.3% 19.5% 39.0% 3.1% 3.0% 15.2%
Kildare South 27.8% 16.9% 49.6% 4.1% 0.0% 1.6%
Laois-Offaly 42.3% 36.7% 7.8% 1.0% 9.2% 3.1%
Limerick City 25.6% 28.4% 31.2% 2.2% 9.5% 3.0%
Limerick 28.1% 50.3% 19.2% 2.3% 0.0% 0.0%
Longford-Westmeath 22.2% 29.8% 41.3% 1.1% 5.0% 0.5%
Louth 25.7% 32.1% 13.2% 5.6% 22.0% 1.3%
Mayo 20.4% 51.0% 21.6% 1.7% 5.4% 0.0%
Meath East 24.5% 26.1% 29.2% 2.1% 5.3% 12.6%
Meath West 33.3% 33.5% 11.3% 2.0% 17.5% 2.4%
Roscommon-South Leitrim 20.2% 32.1% 36.7% 3.2% 7.8% 0.0%
Sligo-North Leitrim 25.6% 38.7% 14.7% 6.5% 13.1% 1.5%
Tipperary North 13.9% 13.6% 23.9% 0.7% 6.6% 41.3%
Tipperary South 22.0% 21.2% 23.8% 1.2% 6.3% 25.6%
Waterford 21.1% 26.3% 29.5% 1.5% 13.1% 8.3%
Wexford 23.5% 31.4% 33.3% 0.8% 9.9% 1.0%
Wicklow 19.8% 21.5% 37.0% 4.6% 6.2% 10.9%
STATE 17.0% 28.0% 32.0% 3.0% 9.0% 11.0%

My guess-timate as to the destination of seats in all of the constituencies (Note: these figures assume that Fianna Fail are guaranteed a seat in Louth because Seamus Kirk, as Ceann Comhairle, will be automatically returned to the Dáil):

  FF FG LB GP SF OTH
Carlow-Kilkenny 2 2 1      
Cavan-Monaghan 2 2     1  
Clare 1 2       1
Cork East 1 1 2      
Cork North Central   1 2   1  
Cork North West 1 2        
Cork South Central 1 2 2      
Cork South West 1 1 1      
Donegal North East 1 1     1  
Donegal South West 1 1     1  
Dublin Central 1   2     1
Dublin Mid West   1 2   1  
Dublin North 1 1 2      
Dublin North Central 1 1 1      
Dublin North East   1 1   1  
Dublin North West     2   1  
Dublin South 1 2 2      
Dublin South Central   1 3   1  
Dublin South East 1 1 2      
Dublin South West   1 2   1  
Dublin West   1 2     1
Dun Laoghaire 1 1 2      
Galway East 2 2        
Galway West 2 1 2      
Kerry North-West Limerick   1 1   1  
Kerry South   1 1     1
Kildare North 1 1 2      
Kildare South 1   2      
Laois-Offaly 3 2        
Limerick City 1 1 2      
Limerick 1 2        
Longford-Westmeath 1 1 2      
Louth 2 2     1  
Mayo 1 3 1      
Meath East 1 1 1      
Meath West 1 2        
Roscommon-South Leitrim 1 1 1      
Sligo-North Leitrim 1 2        
Tipperary North   1 1     1
Tipperary South 1   1     1
Waterford 1 1 2      
Wexford 1 2 2      
Wicklow 1 2 2      
STATE 40 55 54 0 11 6
  24.1% 33.1% 32.5% 0.0% 6.6% 3.6%

Even with a higher share of the national vote than Fine Gael, the model predicts that Labour might not emerge as the largest party in terms of the number of seats won. This is down, to a large degree, to the fact that there are a number of constituencies – particularly in the Border, West and Midlands regions – where Labour would still not be in contention for seats, even if their vote share was to be over three times that of their 2007 support levels, in line with the national trend predicted by the Irish Times/Isbos MRBI poll. Good vote management by Fianna Fail and Fine Gael, in addition to the considerable amount of ‘wasted votes’ by weaker parties in different constituencies (including Labour in constituencies such as Cavan-Monaghan, Clare, Laois-Offaly and Sligo-North Leitrim), would see these parties winning more seats than their share of the vote would warrant  in a number of constituencies.

Advertisements

10 Responses to “Irish Times/Isbos-MRBI Opinion Poll June 11 2010: A Geographical Perspective”

  1. Mr WordPress Says:

    Hi, this is a comment.
    To delete a comment, just log in, and view the posts’ comments, there you will have the option to edit or delete them.

  2. Gavan Reilly Says:

    Adrian,
    Very interesting analysis. I was just wondering what numerical methodology you used – did you inflate each in line with the differences between the 2007 breakdown and today’s polls (allowing for provincial variances mentioned), or is it a bit more nuanced to deal with individual constituency circumstances (such as popular candidates, etc)? If so, which ones proved anomalous to the national pictures?

    • Adrian Kavanagh Says:

      Gavan,
      Party support levels have simply been inflated/deflated in line with changing support levels for their particular provinces; e.g. if Labour support levels double in a region, a 14% support level in a constituency within that province would be expected to translate into a 28% support level based on the poll figures.
      I haven’t taken account of (the infinitely more interesting) local constituency characteristics – ultimately factors such as popular local candidates would change these figures!

  3. neilcaff Says:

    Very interesting. Thanks for this useful analysis.

  4. David Convery Says:

    Have a look at your other for Cork North Central, Dublin West and Dublin North, could this translate as three seats for the Socialist Party? Isn’t it time the Socialist Party was included as a category in its own right?

    • Adrian Kavanagh Says:

      Unfortunately the polls don’t give actual figures for the Socialist Party, so am forced to include this party amongst the Others grouping as a result; nothing else I can do! Though they are Ireland’s sixth largest, I suspect the Socialist Party will need to win at least two seats in the next general election and/or develop a more developed national organisational/support base (remember: the party only ran four candidates in the 2007 election) before the polling companies will start to treat the party as a separate entity on a regular basis.
      The seat in Dublin West in the “Others” column would be a Joe Higgins seat. (He was the only “Other” candidate in Dublin West in 2007).
      In Dublin North in 2007, Clare Daly was just one amongst four SP and Others candidates (winning roughly two-thirds of the votes cast for this category), and Others figure here would include projections for other three candidates also. Looking at her in isolation, I would calculate her figure at 13.5% based on this simulation. Chance of a seat based on those figures, but Labour would be better placed to take two seats.
      In Cork North Central in 2007, Mick Barry was just one amongst six SP and Others candidates (as the second most popular “Other” canddiate, he won roughly one-third of the votes cast for this category), and Others figure here would include projections for other five candidates also. Looking at Barry in isolation, I would calculate his figure at 5.6% based on this simulation. No chance of a seat based on those figures.

  5. dotski Says:

    Adrian,

    A difficulty with your analysis is that the final outcome in votes ends up different to the poll. I’ve crunched those constituency figures, and the composite national vote you are arriving at is;

    FF 21.28%
    FG 27.64%
    LP 28.79%
    GP 3.72%
    SF 10.62%
    OTH 7.95%.

    This i the difficulty with trying to take on regional swings within polls, the various roundings start to accumulate.

    If the poll in question actually came to pass, LP shoul be looking at 62-64 seats.

    • Adrian Kavanagh Says:

      There is always going to be some discrepancy arising when trying to reaggregate up from provincial level figures (especially as provincial figures quoted in article seemed to include significant “Don’t Know” elements; this and the absence of the PDs as a factor now meant constiuency estimates had to be re-standardised so they all added up to 100%).
      Not so sure about your seat estimate for Labour; re-running my analysis just taking account of the national levels figures would leave parties at: FF 34, FG 59, LB 57, GP 0, SF 7, OTH 9. Interesting that the more geographical analysis seems to be giving FF a 6-seat bonus at expense of FG and LB – suggestive at what may lie behind levels of disproportionality within the Irish electoral system.

  6. dotski Says:

    Adrian,

    Those figures would mean the following national figures

    FF 21.28% (4.28% higher than the poll)
    FG 27.64% (OK pretty spot on)
    LP 28.79% (3.21% lower)
    GP 3.72% (bit of a bonus, 0.72%)
    SF 10.62% (1.62% bonus)
    OTH 7.95%.

    I think you have to fine tune as there are rounding issues that you’re not compensating for.

    • Adrian Kavanagh Says:

      As noted above, rounding up/fine tuning was focused at level of generating the constituency estimates which is what the main focus of this exercise was. There is always going to be a discrepancy when standardised constituency data are reaggregated up to the national level.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: