Fianna Fail experienced a significant loss in support in the recent Donegal South West by-election relative to their support level in that constituency, with their support down by 12,792 votes and percentage share of the vote down by 29.2%. But the recent history of the party has shown a tendency for the party to poll poorly in by elections only for party support levels to recover in a subsequent general election. During the lifetime of the present government (1997-2010), eleven by-elections have been held with eight of these taking place prior to the 2007 General Election.
Fianna Fail’s average support level in these constituencies stood at 41.5% (for the 1997-2010 period: support levels stood at 40.2% when jhust looking at the constituencies where by-elections were held over the 1997-2007 period).
Fianna Fail’s average support levels in the subsequent by-elections held in these constituencies stood at 24.8%, marking a decline of 16.7% in the party’s percentage share of the vote. Just looking at constituencies where by-elections were held over the 1997-2007 period, we see that Fianna Fail average by-election support levels in these constituencies stood at 27.8%, marking a decline of 12.4% in the party’s percentage share of the vote in these eight constituencies relative to their support in the earlier general election.
The interesting angle here is to look at Fianna Fail support in the general elections held following a by-election. The trend over the 1997-2007 has been for party support in the constituencies affected to not just return to the levels experienced in the previous general election, but for these support levels to be actually surpassed! In the eight constituencies where by-elections were contested over this period, Fianna Fail’s average support levels stood at 41.5%, marking a 13.6% increase in party support levels relative to their by-election performances but also marking, on average, a slight increase of 1.3% on their support levels in the previous general elections held in those constituencies. Furthermore, the party were to make significant seat gains in a number of these constituencies in the follow-up general election, including Kildare North (2007) and Meath East/Meath West (2007), and always won back any seat lost to the party in a by-election.
These trends are in keeping with the second order election model, which suggest that voter behaviour in elections perceived to be of “less importance” to voters than first-order elections (i.e. general elections) will be noticeably different than that for the first-order contests. By-elections are a classic example of second-order electoral contests…and the recent Donegal South West by-election even more given the particular context that this election took place in. One noticeable aspect is that voter turnout levels will usually be lower in a second order election, and this tends to be the case for by-elections and especially for by-elections fought in urban constituencies. The other major trend that emerges is for support levels for government parties to be lower than those normally registered for those parties in general elections. Second-order elections, such as by-elections, allow an opportunity for government party supporters to register a protest against aspects of their party’s performance in government that they are unhappy with, either by opting not to vote at all or by voting for another party or an independent candidate, before usually returning to their normal voting preferences at the following general election. There is also an artificial dimension to support changes for Fianna Fail, which tends to run two or more candidates in all Dail constituencies, as lower support in a by-election can in part be explained by the party just running one candidate in this instance as opposed to general elections where the party would be running multiple candidates in the same constituency. The fact that popular incumbents would not be contesting by-elections is another factor that can account for the depressed Fianna Fail support levels in by elections. (It is also worth noting that the party’s support levels also declined in by-elections held over the 1995-96 period when the party was not in government, even though they did win two of these by-election contests: Brian Lenihan’s win in Dublin West (1996) and Cecilia Keaveney’s win in Donegal North East (1996).)
What lessons can be gleaned relating to the by-elections held in 2009 and 2010. Trends over the 1990s and 2000s as discussed here suggest that Fianna Fail support will be significantly, or at least somewhat, higher in the constituencies involved in the next (2011) general election. However, the average level of support decline in these constituecies far outstrips the level of declined observed for by-elections carried out during the first two periods during which the current Fianna Fail government has been in power (1997-2007); Fianna Fail support dropped by an average of 12.4% for by-elections held over the 1997-2007 period but dropped by an average of 26.6% for the three by-elections held in 2009 and 2010, the most noticeable drop in support being registered in the 2009 Dublin Central (down by 32.2%) by-election. Such has been the extent of Fianna Fail’s support decline in these constituencies that the prospects of party support rebounding back to, or even surpassing, their 2007 support levels looks highly unlikely.
The basic data underpinning the figures used in this post may be viewed by visiting the new Irish General Election 2011 Facts and Figures blog, or else by visiting the always excellent electionsireland.org website.
Irish Politics Forum 
I think what by-elections tend to show is the low-tide level for any party and the general election shows the potential high tide for some parties depending on whether the tide is in for them or not.
“The trend over the 1997-2007 has been for party support in the constituencies affected to not just return to the levels experienced in the previous general election, but for these support levels to be actually surpassed!”
Of course it has! In 2002-2007, FF support collapsed just after the 2002 election and surged just before the 2007 election. Does your model benchmark with the contemporary national poll numbers to extract the individual effect of by-elections?
Surely we can all agree that if FF support surges just before an election, we’d expect by-election results to be well below general election results.
No we can’t agree on this. FF support doesn’t always drop significantly just after a general election and surge just before the next one – 2002-2007 period was probably an exception that doesn’t prove a rule.
If you study polling figures for when the by-elections were actually held, these will show that FF support levels were actually improving at that point in time following the significant drop in party support in late 2002, ultimately leading up to the poor party performance in the 2004 elections. True, at 37% in the Red C poll of March 2005, FF support was lower than what it would achieve in the 2007 election but it was not significantly lower – the degree of increase in party support in the two by-election constituencies (Meath and Kildare North), at an average of 15.0%, far outstripped this.
By the way, this is not a model. These are statistics of actual voting patterns.
Adrian,
The drop in FF support post the 2002 GE was caused by a series of cuts in public expenditure announced by the government,just a few months after the election and in response to an international downturn. Once the economy improved, public expenditure once again roared forward, albeit financed by the property taxes bonanza from the property bubble that was beginning to take shape. Budget 2007 reads like a fantasy tale from another planet; a pre-election budget that raised pension rates, other SW payments etc. and took a further 85,000 workers out of the tax net plus a little something or other for everybody elese in the national audience. The difference between then and now is that the bubble has spectacularly burst and the state teeters on the edge of sovereign default, even with an IMF/EU rescue package on the table. This after two and half years of cuts and other austerity measures that have already been stomached by the population. In that context I’m not sure how much, if any reliance, can be placed on traditional voting patterns or precedent.
I think you have a more than valid point there Veronica. But I think, and this is the reason behind this post, that it is informative to at least offer the statistics on what the recent (i.e. 1997-2007) precedent has been in order to better grasp/understand what some of the messages are that might be gleaned from recent by-elections.
A 1997-2007 data set is going to be at least partially dependent on the behaviour underlying the 2002-2007 data set… I don’t see why the author chooses to omit an adjustment to account for changes in underlying party support in a post that is, implicitly or explicitly, modelling by-election party support, regardless of whether that word is used.
The prior here seems to be constant party support between the day after a general election and the next general election, which is unlikely to be true, given the significant fluctuations in polling figures between 2002 and 2007. I merely propose that a new, more accurate prior be chosen.
This election is a game changer where all bets are off.
The reality is we have absolutely no idea what will happen to Fianna Fáil in the election because never before have the circumstances been as they are now.
Even during the worst of GUBU and cronyism of Ahern there were still 40% of the population who freely admitted to supporting Fianna Fáil.
That is not the case now and even if there are a few % more who support the party than will admit it to a pollster, will they go and actually vote or just not bother.
Then there will be the determination of all other party supporters to get off their backside and cast their vote and they won’t be giving transfers to Fianna Fáil candidates.
We can play guess games on how badly Fianna Fáil will do based on local, European and by election results. But the fact of the matter is we don’t even know who the candidates still running will be and what their track records will be and it is simply not possible to quantify the effects of an actual general election campaign and whether it makes people more angry or results in one of the opposition making a major mistake.
As interesting as it is to guess, we simply have no way of knowing.
I think you make very valid points Desmond. In light of the recent departures, it would seem presumptuous, to say the least, to assume that Fianna Fail will even contest every seat. It may be the case that they choose to baton down the hatches, close rank and defend their remaining strongholds at all costs. They may see it as an opportunity to re-invent the tarnished brand that is Fianna Fail, which may seem un-salvageable now, but may appear totally different in ten years. Honestly, I do not think their is any way back from this, but time and political failings by the opposition may alter that perception over the next decade. One only has to look to America, and the destroyed Republican party shell that was left following the last election. Having led the country into two wars, pushed the nations debt into astronomical figures and destroyed the country’s reputation as a democratic state, it seemed like the party was facing into at least two terms in opposition. Now, two years later, the Republican brand has re-emerged stronger than ever, with a rejuvenated base and an almost fanatical following in certain quarters. Obviously, the Irish situation is much different, as are the Irish people, whose memories are not so easily washed clean, but it is definitely food for thought…
While it was clear in previous by-elections that the government would lose, it was never so blatantly apparent that the government has lost the confidence of the vast majority of this states citizenship. This was a direct vote on the Fianna Fail party’s ability to rule, regardless of the spin they have since tried to put on it, and the people spoke out with a clear response. It was clear by the time the first count was in that never since the foundation of this state had an administration managed to polarize the people in such a way. I have to admit, as a former Fianna Fail voter myself, that I personally feel betrayed beyond belief by this government. They have, to use the vernacular, sold us down the swanny and they do not even have the decency to allow our voices to be heard. We need an election immediately and we need our government to put this settlement before the Dail for ratification. Surely, if history has taught us anything, it is the danger that occurs when you marginalize the mass of the people in favor of an elite few? It is a recipe for revolutionary upheaval. The people are enraged at the bankers, who have slipped away in to the shadows, and amazed at the cheek of the Union leaders to call for the mobilization of the poor, while they sit in their mansions and refuse to take substantial salary reductions. Do we really believe a Union leader should be paid two hundred thousand euro a year? Or, on a different, but undoubtedly connected, note that the head of the ESB is worth seven hundred and fifty thousand euro a year? The combination of hypocrisy of pure cheek that permeates this states ruling bodies truly beggars belief. Where is the leadership we all so badly need at the moment? Surely someone will stand up and be counted? Perhaps Eamonn Gilmore, the man who would undoubtedly make a better leader than the media dodging Michael-Davitt-inspired Kenny, should rise the people up in a move toward change? He is articulate and quietly spoken, two qualities that are a necessity for modern political leadership, or so it seems. However, he is in real danger of running weak candidates at such an important time in this states history. His choice for Donegal, Mr McBrearty, was clearly a poor choice for the party to make, in such a publicly scrutinized race. His arrogance during a Radio one interview seemed palpable, as did his inept ability to justify his position regarding Sinn Fein and his seemingly endless desire to answer a question with a quesion. He really came across really poorly. One can only hope that Mr.Gilmore learns from his mistake and does not, as countless Labour governments have done throughout the states history, fluff his opportunity at the last minute.
The analysis was somewhat odd. It has validity in arguing that by-elections are not a valid means for predicting general election outcomes, with Fianna Fáil in particular doing far better in general elections than by-elections. It then however contradicts all that in the coda in the interpretation of what would be the outcome of Labour if Labour would perform as in that by-election.
That analysis is flawed. The trebling of Labour support was illusory – the party did so badly in 2007 there that it was always going to see an increase that would be a multiple of the 2007 vote.
In practice Labour is the party with most to fear from what happened in the by-election, simply because it matches a trend – of Labour consistently in a range of elections doing less well than polls suggest. One can argue, correctly, that by-elections are not a reliable guide to general elections. What is impossible to avoid is the fact that in yet another type of election Labour came in significantly below what polls said. That must for Labour raise the fundamental worry that polls that show Labour in the 20s or higher may be significant overestimates of the real Labour support in an electoral contest.
If so, and Labour runs candidate numbers based on national polls, Labour could end up losing seats as it did in 1969 that would have been won with less candidates and so less split votes in some constituencies.
Another serious worry for Labour must be candidate selection. Labour has followed a strategy of choosing local notables, often prominent figures or ex-TDs who belonged to other parties or were independents. The selection of McBrearty went disastrously wrong – he bombed on the doorsteps, got into rows with voters, and in at least one stage with a guard, and made a mess of his media appearances. None of the other new Labour candidates are likely to be as bad, but the worry must be – ‘have we been choosing the wrong type of candidates?’ Cowley, Sexton, etc are not exactly highly popular figures, for example. (Other parties were astonished Labour touched them with a barge pole.)
Organisationally, also, something went wrong when a party with the resources of a mainstream national party, able to flood a constituency with Oireachtas members and canvassers from all around the country, gets beaten by an independent with a minimal support base and tiny organisation. For a major party with a bused in organisation (all parties can do that in by-election) to be beaten by an independent, when that independent has not established themselves as the ‘lead’ challenger, is frankly astonishing.
Finally, the success of Sinn Féin gives Labour more than any other party something to worry about. Fianna Fáil initially lost some support to Fine Gael early in this Dáil. Fine Gael then lost some support to Labour, which also picked up other FF support. Labour in effect outflanked FG, where FG previously had outflanked FF. Does the success of Sinn Féin mean that public opinion is on the move again, with SF now possibly going to outflank Labour by eating into the support it has gained? In other words, could Sinn Féin to Labour what Labour earlier in the year did to Fine Gael, and what Fine Gael in 2009 did to Fianna Fáil, and steal its thunder, tapping into the public mood?
Is it the case that, as the shape of the crisis developed, could the public mood be becoming more radicalised, moving constantly from one party to a more radical alternative?
In conclusion, of all the parties in the campaign, the one with most to worry about, in my view, is Labour. Yet again it got significantly less support than the polls suggested. Its tripling of the vote was meaningless given its extremely low base. Its candidate selection failed. It was outflanked as the voice of radicalism by Sinn Féin and it even got beaten, despite the fact that it had far more resources, than a low-key independent. Fine Gael and Fianna Fáil can point out to a distinction between their by-election and general election performances. Labour don’t have that option because the basic problem, of doing far worse than polls suggested, is cropping up across a range of different types of election, and not just by-elections.
Agree with you Jim – “coda” is out of kilter with general theme of post, unbalancing it and is rather clumsily included at end; I will edit the post and I’ll detach from main article and include as an individual comment on this post or the earlier DSW post. Main point here was to try and offer a quirky take to show that while Labour did not poll as well as some were expecting, result did mark a significant increase in party fortunes in this constituency which would be one of party’s weakest bases of support in the constituency!
Here is the “coda” which originally formed part of original post!
Coda: Although some analysts suggest the 2010 Donegal South West by-election result was disappointing for Labour, if the degree of party support change registered for this contest was replicated across all constituencies in the upcoming general election, then Labour would emerge as the strongest party nationally! Using a similar methodology for my previous poll analyses (and holding Green and Independent support levels as constant (as Green Party 2010 figures to compare with in Donegal South West case and 2007 Independent figure not specifically comparable), Labour support would be estimated at 35.7%, Fine Gael support at 22.3%, Fianna Fail support at 17.6% and Sinn Fein support at 13.1% – this would leave the tally of Labour seats in the low 70s, allowing the party to enter coalition with Fine Gael or Fianna Fail or Sinn Fein.
Surely what stands out for Labour is that a popular leader is insufficient to capture seats and what will matter is who the name on the ballot is. In that sense the capture of Cowley and Davy Fitz in Clare might point to Labour being aware of this problem already and seeking to address it with names that people are familiar with. Again a name alone will not be sufficient but it will help them.