Wednesday 27 May 2009

Increase in the electorate – who gains?

The SDLP has received some research from American researcher Dr Michael Moriarty (quoted in the Irish News) which suggests that they could benefit from the recent increase in the electoral register. Ronan McCay, SDLP Director of Organisation and Election Planning, writes that this research implies that:

" … we’re very much in the race. The Irish News article shows that this election is game on, that it’s in our hands and that it can be done."

Dr Moriarty's findings are, as yet, not in the public domain. But the parts quoted in the Irish News and on the SDLP Youth website are simply erroneous.

The article states that:

"Research for the SDLP by an American, Dr Michael Moriarty, points to 65 000 potential new voters on the register and claims that the breakdown is approximately 50 000 nationalist and 15 000 unionist. He adds that the party needs an additional 20 000 votes to win a seat."

However, a very quick consultation of the statistics published by the Electoral Office shows that, since the most recent election (the 2007 Assembly election) the electorate has only increased by 49,949. Even if Dr Moriarty is using the electoral statistics from the last European Parliament election in 2004 he is wrong – the increase since then has been 85,184.

Dr Moriarty's methodology for estimating the breakdown of the increase is unknown. But it is an established fact that deaths since 2007 have been around 30,000, and around two-thirds of these have been Protestants (and thus presumably unionists). New entrants to the electorate (those reaching their 18th birthday) have been around 50,000, of which slightly more are Catholic than Protestant. Putting these two figures together it is possible to estimate that the 'natural increase' in the electorate since 2007 has been around 20,000. Of these, largely due to the Protestant majority amongst the elderly, around 17,000 are Catholic and 3,000 are Protestant. So the net natural increase of Catholics in the electorate is less than 15,000.

Yet the electorate increased by 49,949 during the period. Evidently the Electoral Office has improved their registration rates, and it has to be assumed that these extra registrations are more or less representative of the areas in which they live. Multiplying the increase by the known voting breakdown (unionist or nationalist) in 2007 allows us to estimate how many of these new registrations were probably unionist and how many were nationalist.

Unfortunately for the SDLP (and nationalism as a whole) the increase in registrations was slightly stronger in unionist areas, so the effect of the improvement has marginally helped unionism.

Of the approximately 50,000 new voters in 2009, 20,000 are due to 'natural increase', and 30,000 are due to improvements in the register.

The 20,000 (natural increase) break down as 17,000 nationalists and 3,000 unionists.
The 30,000 (additional registrations)break down as 14,000 nationalists and 16,000 unionists.

The total increases in the potential voters for the two tribes is, therefore:
- Nationalism – plus 31,000
- Unionism – plus 19,000
Thus giving nationalism a net increase of 12,000 since 2007.

But turnout rates in the European elections are low (51.7% in 2004), and there is no reason to think that this election will buck that trend in any significant way. So the actual net increase in the nationalist vote may be barely 6,000 – shared between Sinn Féin and the SDLP.

If the SDLP needs 20,000 extra votes to win the third seat, it will not succeed.

16 comments:

hoboroad said...

Thanks for clearing that up Horseman.The SDLP should put up or shut up trying to decieve the nationalist electorate could rebound badly on them.

Pierre Brasfort said...

I met Dr Moriarty when he was in Belfast. He's a professional statistician. I'm sure his methodology was a bit more advanced than your "very quick consultation"!

hoboroad said...

Well there are lies damn lies and statistics!

hoboroad said...

I would be happy to be proofed wrong on this one but I think Horseman has got it right.The SDLP should publish or be damned.

hoboroad said...

Every Nationalist should vote on Thursday June 4th 2009 make every vote count!

Faha said...

Horseman,
I will clear up the confusion on these statistics. Part of the problem is that people are comparing different elections and different time frames. First we must start with the 2004 European election.
The first count total unionist vote was 266,925 and the nationalist vote was 232,100 for a unionist nationalist gap of 34,825.The final count showed a total unionist vote of 284,378 and a total nationalist vote of 253,072 for a unionist nationalist gap of 31.306.
There have been changes to the voting age population in the previous 5 years. There are new voters ( those who have reached the asge of 18 ) and voters who have died. If you look at the 2001 census the age cohort aged 10 through 14 were not old enough to vote in 2004. However, in 2009 they are now old enough to vote. That age cohort consisted of 66,875Catholics, 60,132 Protestants and 5,657 None/Other. I assumed that the None/Other group would vote 2 to 1 unionist. This gives a total of 68,761 new nationalist voters and 63,903 new unionist voters. However, during the previous 5 years 72,500 voters have died. These would be mainly elderly. This group is 70% unionist in the 80+ age cohorts and 67% unioinst in the 70 to 79 age cohort. Since there are also some younger voters who have died I assumed the overall breakdown was 66% unionsit and 34% nationalist. This translates into the death of 24,500 nationaslist voters and 48,000 unioinst voters. The net overall change for all voters is approximately 44,000 new nationalist voters and 16,000 new unionist voters. Of course, only 50% of these will actually vote in a European election. This would result in 22,000 additional nationalist votes and 8,000 new unionst votes for a new gain of 14,000 nationalist votes. Subtracting this net 14,000 from the 31,306 unionist nationalist gap in 2004 results in a unionist nationalist gap of 17,306 in 2009.The 20,000 figure quoted in various sources is based on a lower turnout than 50% of the net new voters.
As far as the those on the electoral register are concerned there were 1,076,000 on the June 2004 register and 1,158,000 on the May 2009 register for a difference of 82,000. Assuming that those net 60,000 new voters register at the same rate as the remainder of the population then 52,000 of these would be on the May 2009 register. There are an additional 30,000 other new voters on the register. However, the nature of these new voters are not what most people believe.NISRA has recorded 10,000 applications from Republic of Ireland immigrants in the previous 5years. This would be a minimum of those actually moving to NorthernIreland since non working spouses orpartnersand the self employed would not show up in thesenumbers. Even assuming only 10,000 would mean that perhaps 8,000 of those 32,000 are from the ROI. I doubt that any will vote for unionist parties. There is a separate register for Westminsterelections since EU nationals are not allowed to vote in Westminster elections. For May 2009 , there appear to be 8,250 EU nationals on the elctorate. Since these people did not begin arrivng until 2004 almost all these EU nationals are new to the register since 2004. In addition there are an unknown number of new voters from Commonwealth nations of Africa, Asia, Australia, etc who are new to the register. Thus, the majority of these 32,000 additional new voters are not native to Northern Ireland.
Thesenew ROI and foreign national voters are unlikely to vote for unionist parties so the SDLP probably needs 10,000 stay at home voters to win the seat. An uphill battle, yes, but certainly feasible.
Of course, this analysis is assuming that all other factors are equal, which they are not. In the end , I believe it will come down to how the transfers are distributed.

Dazzler said...

I understand why people from ROI would only vote nationalist, but why are foreign nationals unlikely to vote unionist. Are you saying for example that poles will vote sdlp or sinn féin as they are catholic?

Horseman said...

Faha,

Thanks for taking the trouble to explain it to us. I can't comment in detail at present, but will try to respond tomorrow.

BillyM said...

Dazzler,
My guess is that it would not take an Eastern European Catholic with reasonable psychological antennae to pick up on the anti-catholic understreams within unionism and to absorb certain messages from recent attacks on Poles for example. As the tribal breakdown in NI is defined by religion my guess is that assimilation will definitively involve their following religiopolitical suit. Put simply they would feel more comfortable mixing in Catholics circles.

Anonymous said...

We should also remember not all are mew as the electoral registers were decimated by an over zelous electoral office over previous years that put people off the registers for ridiculous reasons.

Anonymous said...

new

Horseman said...

Faha,

I think that on most things we are generally in agreement, except for two issues, though; see below (I have to cut some text to fit within space limitations):

You: "There have been changes to the voting age population in the previous 5 years. …"

Me: Agreed, up to a point. Don't forget the large out-flow at ages 18-19. I think your figures overstate the number of new entrants to the electorate.

On deaths, I agree with you.

You: "The net overall change for all voters is approximately 44,000 new nationalist voters and 16,000 new unionist voters. Of course, only 50% of these will actually vote in a European election."

Me: But, unfortunately for the purity of your maths, the electorate figures do not show 60,000 additional voters since 2004. The 2004 electorate was 1,072,669, and the latest 2009 electorate (for the EP election) was 1,157,853 – an increase of 85,184.

You: "This would result in 22,000 additional nationalist votes …"

Me: "Our analyses reach almost the same figure by different routes – assuming a 50% turnout I reckoned that since 2007 nationalism would gain 6,000 actual net votes. If you extrapolate this back to 2004 (5 years instead of 2) it comes to approximately 15,000 net new votes. Still not yet enough to bridge the gap of 31,306."

You: "As far as those on the electoral register are concerned …"

Me: As stated above, I think you overestimate the number of new entrants to the electorate at age 18. I think it is closer to 25,000 per year, giving 125,000 over the period 2004-2009. With deaths at 72,500 (I agree) the 'natural increase' is 52,500. Unlike you, I think that most resident adults are on the register.

You: "There are an additional 30,000 other new voters on the register. However, the nature of these new voters are not what most people believe …"

Me: Well, NISRA use registration with a doctor to count migration, so their figures probably cover most migrants. They may include some fake migrants, as people living just south of the border may use a doctor in the north because it's cheaper.

You: "There is a separate register for Westminster elections since EU nationals are not allowed to vote in Westminster elections. For May 2009, there appear to be 8,250 EU nationals on the electorate. Since these people did not begin arriving until 2004 almost all these EU nationals are new to the register since 2004. In addition there are an unknown number of new voters from Commonwealth nations of Africa, Asia, Australia, etc who are new to the register. Thus, the majority of these 32,000 additional new voters are not native to Northern Ireland.
These new ROI and foreign national voters are unlikely to vote for unionist parties so the SDLP probably needs 10,000 stay at home voters to win the seat. An uphill battle, yes, but certainly feasible.
"

Me: I think that you are forgetting one very important point. The 30,000 'new registrations' are not all migrants. They are, in most cases, people who were removed or excluded from the register previously, during the Electoral Office's draconian period. I have 'replaced' them as if they are representative of the areas in which they live, and a slight majority are likely to be unionists. So you are perhaps underestimating the unionist vote, and thereby underestimating the hill that nationalism needs to climb.

You: "Of course, this analysis is assuming that all other factors are equal, which they are not. In the end, I believe it will come down to how the transfers are distributed."

Me: I agree, but I think we all know that unionists will transfer to unionists, and the 'other' will probably break fairly evenly, so first prefs are still very important, and I just don't see nationalism getting enough, unless there is a lower unionist turnout.

Faha said...

Horseman,
Thank you for your reply. I do believe that most of the 60,000 reaching voting age is the actual number. NISRA estimated a net out migration from 2001 ( census date ) to 2004 of only -1500. Since 2004 they estimate a net in migration of 22,000. However, 2004 is the year the large EU in migration begun. NISRA does mention the large out migration that occurs from age 18 to 22 ( probably mainly students who go to the ROI, England and Scotland ). However, they also document a gradual return of those who leave. In the 2001 census there were 26,276 age 17, 25,392 age 18 and this declines to a nadir of 21,116 at age 23. However, this number gradually recovers so that it was 25,275 at age 30 and 26,845 at age 36. Since the original birth cohorts for these ages were almost identical this accounts for the fact that since 2001 there is little net out migration of NI natives, What little net out migration is likely to be unionist students who go to school in the UK and do not return. You have previously documented this.The Northern Ireland Electoral Commission site has a report on the electoral register. The registration rate of the adult population is only 84%, though higher now this close to the election. That is why I estimated that 52,000 of the net 60,000 of voting age since 2004 are on the register. Yes, since there are 85,000 more people on the register since 2004 then there must be an additional 33,000 voters that are not due to demographic changes. However, I still believe that the majority of these are not NI natives. There have been over 10,000 NiNO applications from immigrants from the ROI over the previous 5 years. In addition there would be some non working spouses and partners who do not apply for NiNO. So I estimate that at least 8,000 of those 33,000 are from the ROI. The 8,000 EU nationals that are new to the register since 2004 must also be accounted for plus an unknown number of other foreign nationals.
You metnioned that there appears to be a greater increase in registration in unionist areas.However, at least in Belfast this does not appear to be so. From June 2004 to May 2009, the electorate in East Belfast increased from 49894 to 51,015 (+1121), in South Belfast from 49,473 to 50,795 (+1322), in North Belfast from 49,576 to 52,114 (+2538) and in West Belfast from 48,869 to 53,013 (+4144). It is obvious that the greater the nationalist population in a constituency the greater the number of new voters on the register. Lagan Valley had an increase of 7,902. However, 2,229 of these new voters are in only 2 wards, the heavily and inreasingly nationalist Derryaghy and Glenavy wards.
I believe the SDLP could win a seat. However, it will require 10,000 stay at home voters. You mentioned that unionist voters transfer to unionist parties. However, Alliance and Green voters usually transfer to the UUP or SDLP. If Sylvia Hermon is unwilling to support the UUP-Conservatives then how willing are Alliance and Green voters to transfer to the UUP-Conservatives ? We will find out in 1 week !

hoboroad said...

Big debate on the BBC1 at noon.All the ones who want your vote on Thursday will be there.

hoboroad said...

Cannot wait till to Thursday

Anonymous said...

Tá an lá ag teacht