The same point keeps coming up in comments – if there is some great surge of young voters backing the Liberal Democrats, would it be picked up in the polls?

The short answer is that it should be. Pollsters in the UK do not weight their samples to match the demographic profile of voters, they weight them to match the demographic profiles of UK adults, so most of those young people who wouldn’t normally have voted should have been represented in the samples anyway. Only Harris specifically ask respondents if they are registered to vote, so they would not have been filtered out of other companies’ samples.

Another point I’ve seen raised is how well pollsters cope with young people who are predominantly online or have only mobile phones. Firstly, online really isn’t a problem, since we have plenty of online pollsters. Secondly, all polls weight by age, so this cannot result in an under-representation of young people. The problem would be if mobile-only young people were significantly different to young people with land lines. If landline penetration continues to fall I suspect that this will eventually be a problem that phone pollsters need to tackle, though there is always the option of including mobile phones in samples.

Perhaps the most common question I’ve seen is whether a change would upset pollsters “assumptions”. Pollsters do not assume particular groups are more or less likely to vote. Instead the majority of pollsters factor in likelihood to vote by asking people how likely they are to vote, and then weighting or filtering appropriately. The correct proportion of young people are already represented in pollsters samples, so if those people told pollsters they had become more likely to vote, it would be picked up.

The bottom line on a lot of questions of whether pollsters would pick up a new trend is that pollsters don’t actually make many presumptions up front about how people will behave. With a few minor and well evidenced exceptions (such as Populus and ICM’s reallocation of don’t knows on the assumption they are likely to vote how they did last time), voting intention figures are based on what people tell pollsters, not the pollsters’ preconceived assumptions.

Things can go always go wrong of course, and I expect it’s more likely to happen at an election where there has been a large shift in support, but I can’t see any particular reason to expect the polls to get it wrong this time.


104 Responses to “Young people and the Lib Dems”

1 2 3
  1. Further to Rashid’s response to Geoff.

    I can’t help but feel that there is a potential circularity in the basis for weighting poll responses. As Rashid says, It assumes no systematic difference between those polled and those omitted e.g. because they don’t have home phones, or use the Internet. However, there would need to be some external validation of the absence of systematic differences (e.g. a polling company using different selection criteria that would include those with mobile phones only/face-toface polling). Without such external validation the results would be self-fulfilling.

    So this leads me to ask someone who knows what they’re talkign about (Anthony?) what would have actually happened with the LibDem surge? Wouldn’t the pollsters, when first picking up an 8% increase in the LD vote, have concluded that their samples were over-weighted to the LDs and therefore would have scaled back the increase? How did they distinguish between an untypical sample and a real change in the position of the LDs?

  2. @SueMarsh

    I see the Mail have an article on how a hung parliament would turn Britain into Italy politically – your dream come true!

    And talking of Europe, I have also seen how Zapatero in Spain is being cold-shouldered by Sarkozy, Merkl etc. (France Soir 14 March). An example of what happens to a lightweight leader.

  3. @RogerH

    Most commentators seem confident it will be higher than 2005.

    Lots of comment too about lack of interest, low number of posters out and about, a large number of undecided according to some reoprts, something the polls should reflect IMO.

    So I think my prediction is on safe ground, but I will of course admit defeat if the election proves me wrong

  4. Can someone explain the Sporting Index (implied) figures on the Electoral Calculus list of polls?

    Sporting Index (implied) 24 Apr 10 – 24 Apr 10 10,000 36.7 27.4 25.6

    I simply can’t understand where they get that number from.

  5. Interesting to see how little impact ‘others’ seem to be having in the media (and in the polls). I think with the long term decline in party brand loyalty (shown, most noticeably in the LF adding 50% to their vote in the polls) that ‘others’ (excluding Nationalists) may be in trouble.

  6. ‘And talking of Europe, I have also seen how Zapatero in Spain is being cold-shouldered by Sarkozy, Merkl etc. (France Soir 14 March). An example of what happens to a lightweight leader.’

    France and Germany have always ignored Spain; and the rest of Europe. Check your history Greengrass.

  7. @MikeP, Xiby

    Purely from an emprical perspective, judge the authority of one source’s predictions by its provenance. Either of the two sources cited could be forecasting a 200+ seat victory for whatever party and it would not fidge me.

    ICM and YG and in fairness to ComR are tested and reputable companies. i would not worry yoursleves with roving mavericks.

  8. Further to Cliff’s question, what is the latest national % for “likely or certain to vote Don’t Knows” in the various recent polls please? I do not find the way that it is excluded from the officially published results helpful. Does anyone else agree?

  9. DC playing table tennis… Saying too many oopses and sorries, and pretty much making a fool of himself.

    Its simply cringe. Not sure its such a bad stunt though.

  10. @JACK

    ‘Interesting to see how little impact ‘others’ seem to be having in the media (and in the polls). I think with the long term decline in party brand loyalty (shown, most noticeably in the LF adding 50% to their vote in the polls) that ‘others’ (excluding Nationalists) may be in trouble.’

    Others are on 10% which is about 2m of the 20m who vote,that could be a lot of seats effected if they are being used in a more targetted way than in previouis elections, and believe nme if the BNPonly finish 2nd in Barking and Nigel Farrage goes close against Bercow and the Grrens win or go close 2nd in Brighton it will be an impact

    remember the Euro elections, BNP wins two seats was the only story in town

  11. Leslie – you can’t weight by current voting intention, or it would indeed be circular (whatever people said, you would always weight it to your own result!).

    For pollsters who use political weighting, they need to weight people’s political opinions at a fixed moment in the past. In most cases this is how they voted in 2005. A representative sample of the population will always have cast their votes in 2005 in the same way, however their political opinions have changed since then.

    (Actually it’s a bit more complicated than that, because recall of past vote does change slightly over time due to false recall, but you get the picture)

  12. Will be be getting polls from Angus and populus?

  13. @Eoin

    ‘The last minute surge in registration of 100,000 voters added c.150 voters per constituency’.

    But if it is only say 50 university orientated constituencies that the bulk of these new registrations are mainly confined to, then that make the new registrations up towards 2,000 in these specific constituencies. That must surely be significant. In Newcastle East only 1500 new LD votes would represent a size-able chunk of the Lab majority.

  14. ‘remember the Euro elections, BNP wins two seats was the only story in town’ basically they won due to low turnout of course.

    Others 10% but take away the Nationalist parties… What are the exact figures for Nationalists?… I suspect the vast majority. That leaves the normal few % points for the extremists and loonies which will be swamped. Others were 12% 30 March on Yougov poll. This is what I mean about them being squeezed.

    BNP and UKIP may come 2nd; but that’ll be loser land behind what they thought they’d get. The fringes thought they’d pick up the ‘antipathy’ to the big parties vote which have gone to LD.

  15. @ Anthony Wells
    ‘Pollsters weight to the profile of the adult population, not the profile of the registered adult population. They were always included in pollsters samples.’

    Surely the polls are not weighted to include the whole adult population as if all were registered and eligible to vote. Surely there must be a weighting factor figure that assumes a quota of the sample will be ineligible to vote.

  16. And just consider the media inches the LDs have gained; because it is now a three way race the fringes are not getting any media miles…

  17. @Xiby:

    See h t t p://www.electoralcalculus.co.uk/campaign2010.html for an explanation of this “Sporting Index” stuff.
    Scroll down to “2. Spread Betting Markets”

  18. Anthony

    To be inappropriate to the title of the thread (as usual) can I ask about older voters?

    You said that YouGov polling rests on the assumption that, within a certain demographic group, online users are politically representative of those not connected. I can see this being true of younger age groups up to say 50 – 60, but those older than that may not be the same, especially those whose children left home before the Internet arrived.

    In this group, overwhelmingly retired, the cost of connection/use and the unwillingness to learn new skills, may make many unwilling to use the Internet. Those that do are more likely to be professionals, the AB social groups and so on.

    This in turn may make Labour voters under-represented among you respondents. Is allowance made for this?

  19. @Frank:

    I think the pollsters assume that pretty much everyone who is not registered to vote is aware of this and will therefore respond “I’m not voting” to the VI question.

    It’s a fair question as to whether the group of people who think they will vote, but won’t be able to because they haven’t registered, will be large enough to have a significant impact on the result this time.

  20. @FrankG,

    c.1500 and that would be a very big if. In addition, they often vote back in mummy and daddy’s constituency.

  21. @Roger Mexico………….I’ve just been down Jamaica Rd on the C10 bus, the tower blocks are still there but Simon doesn’t seem to have worked his magic yet. Millwall George Cross flags in abundance, funnily enough, Simon’s constituency HQ is in the Blue, a market square in Bermondsey, the local market traders generally kick off a conversation with, ” I’m not a racist, but…………!” Mind you they love Si. :-)

  22. @Eoin

    You clearly have a lot of knowledge and have spent a lot of time working on figures. I enjoy reading your posts.
    But. No disrespect, but the polls are not in line with your predictions, and don’t seem to be moving more towards them. What is your confidence level in your predictions, and what do you base that on?

  23. @Jack
    “France and Germany have always ignored Spain; and the rest of Europe. Check your history Greengrass.”


    Are we talking Guernica here?

  24. @MrsB,

    Thank you.

    Blue started 21 days ago with an 8.9% lead in the polls. With YG , ComR and ICm last night they had a 4.66% lead. Labour can finish 2% behind and still form the next government WITHOUT the support of LIberal Democrats. A bit more tightening of the polls and polling booth jjitters and I won tbe too far away ;)

  25. @GreenG

    Guernica is not Spain ;)

  26. @ Eoin

    “In addition, they often vote back in mummy and daddy’s constituency”

    Any evidence for that? I’d replace ‘often’ with ‘rarely’. Those students that do vote will, I expect, vote in their university constituency.

  27. Looking at the BBC’s article on the size of cuts post election their Eco Editor says:

    ” The Conservatives would be looking for total tax increases of £14bn, including many of those proposed by Labour. But the pledge to avoid Labour’s National Insurance rise, along with other Conservative promises, will probably mean the Tories would need to raise taxes by an extra £3.5bn over the Parliament. They have not indicated whether, or how, they would do this”.

    It might, therefore, be deduced that VAT (as well predicted by LDs) may well increase substantially. As VAT has only ever been increased by Tories – if this revenue raising scheme is debated by the leaders on Thursday, how might it affect the post debate polls?

  28. In a two and a half seat council I know it was 1600 late registrations. Has a sudent base but not an absurd one.

  29. @EoinClarke
    “Guernica is not Spain”


    :-)
    I chose the place advertently, as France was also mentioned by Jack. You know that there is DNA and other evidence of post-Ice Age (proto-Celtic) migration from Basque country along the Atlantic to our Isles?

  30. Should have added that I’m thinking of the ‘young people’ who predominantly like luxury products (mobiles, etc) and if VAT goes high on these goods and given that unemployment is high amongst young people – how might they react ?

  31. @Pankot

    it has to be England to find places where electoral predictions are based on betting markets.

    Simply inconceivable.

  32. @RogerH

    Plenty of evidence. Fermanagh and South Tyone, West Tyrone, Newry and South Armagh, Foyle, all are areas of nationalist support. The big city of Belfast has two Universities. And a further education college…. all together 60,000 students….

    The Catholic students vote at home. The only republican candidate in teh Uni area polls c.3000 votes and that is from the Lower Ormeau, Markets area of Belfast.

    I hope this helps :)

  33. Holding back the flood ain’t easy, FTSE serious losses today, Greek bonds marked down to junk, IFS playing devil’s advocate on tax and spend, serious holes in revenue, who would want to win an election ? :-)

  34. @GreenG, :)

    I have many friends there. One of my v.good friends wrote a book comparing the two ethnicities. Very interesting indeed.

  35. Eoin
    Like MrsB I enjoy your well argued posts very much indeed. I can see that keeping your eye on the Con/Lab gap has led to movements that are in line with your predictions. However, I do feel your prediction that the LDs will finish up a shade lower than 2005 highly unlikely. I think from here whether they go up or down depends largely on Thursday night.
    Historically, as ‘likely and certain to vote’ DKs come off the fence, the LDs normally go up in the last few days – yet you predict a sort of meltdown back to their 2005 level? Why? I agree with you that Con and Lab will be tight, but I expect LDs to hang on in there probably – but of course, fayre worst in seats.

  36. @ Eoin

    “I hope this helps”

    Not really. What relevance does the preference of NI’s RC students have to elsewhere in the UK?

  37. @Jack

    Others are on 10% which is about 2m of the 20m who vote,that could be a lot of seats effected

    Others at GE2005 was 7.99 of the population. The notional figure being used a s the baselin will have changed fractionally.

    In 2005 the breakdown of Others included

    Total Others vote including SNP & PC

    England 1,353,133

    Scotland 101,754 plus
    412,267 SNP

    Wales 68,981 plus
    174,838 PC

    Total Others 2,110,953 = 7.99% of total UK vote

    Most of these are for SNP,PC,UKIP,Greens, BNP and a plethora of indepandant candidates of which some are already MPs.

    I do not think that an estimate of 10% for Others, given the disillusionment of the electorate is an overstatement of their %. Indeed I would think it could be slightly higher. Don’t confuse Others with Don’t know or Don’t care. These a often dedicated supporters and their likelyhood to vote coould be very high.

  38. @RogerH,

    Well they are fellow students…In addition the protestant students do the same for East Antrim etc… So it may be a student phenomenon replicated throughout.

    Do you have any evidence to the contrary?

  39. Anthony, thanks for your response to my query about weighting. I’m still unclear though as to how the pollsters distinguished between an LD surge and an unbalanced sample, if they weight back to 2005 results when the LDs got 22% (or indeed to any a priori date). Specifically, when obtaining a 30% figure how did they identify this as as a real shift rather than as evidence of an unbalanced sample that needed weighting back towards 2005 numbers?

  40. @Tony D,

    The best thing about this is that we will not have long until we know one way or the other. Cosnidering the hit 34 and some have them hitting 28 and that is just in one week. The same just need happen this week for that to be replicated.

  41. Tony,

    I hope you will note that I did not use the 26% and 27% from ComRR since we resolved that was an outlier

    I hope you will note that I did not use the 23% from MORI, since we resolved that that was an outlier

    The 34% was fully believable at the time.

    Remember, LD led or joint led in 6 polls last week

  42. If my kids are any indication, they don’t/didn’t use the landline, when living with me. They wouldn’t answer the phone as all their friends ring their mobile, so I would say you wouldn’t get kids living with their parents by ringing landlines.

    BTW why is Guernica not Spain? It wasn’t bombed for being Basque, it was for being left-wing.

  43. @EoinClarke

    I lived through a time in Madrid when Basques were reluctant to drive there with Bilbao or SS registrations and could be thrown into the police cells simply by reason of their surname. Some others were thrown out of a window in the Bohemian tradition, according to the grapevine at the time.

  44. @ Roger H
    “Any evidence for that? I’d replace ‘often’ with ‘rarely’. Those students that do vote will, I expect, vote in their university constituency.”

    First year students will generally be registered in their parents constituency as they will still have been living at home when they turned 18.

  45. @ Eoin

    “Well they are fellow students…In addition the protestant students do the same for East Antrim etc… So it may be a student phenomenon replicated throughout.”

    Since many seats divide on nationalist/loyalist lines it’s not surprising that NI students might return to their ‘home’ constituencies to vote. The circumstances don’t exist outside NI.

    “Do you have any evidence to the contrary?”

    If more students vote this time it’ll be because of the campaigning and mood they’ve experienced. The surge in late registrations – which the Electoral Commission says is 40% from the 18-24 age group – has taken place during term time. Fewer voting students are likely to be registered at their parents’ address than at their university address.

  46. @RogerH

    considering the Math,

    If Youth is 40% of 300,000 = 120,000

    that is 159 votes per constituency.

    And I even excluded those pesky Northern Irish ones for you.

  47. @Roger H
    @ Eoin

    “In addition, they often vote back in mummy and daddy’s constituency”

    Any evidence for that? I’d replace ‘often’ with ‘rarely’. Those students that do vote will, I expect, vote in their university constituency.

    In which case, back to what I have been trying to say for some time. If the LD student population votes in the university constituencies then what may seem a surge will feel like a sunami. Hence the concern about such seats as Newcastle’s. Conversely the LD surge will be less than expected outside of the Unis, especially if those student who would normally vote in their home constituencies vote away from home.

    Remember also the students have been registering for some time, these must be added to the numbers not just the last minute registrations.

  48. @DerekPierson
    “is Guernica not Spain? It wasn’t bombed for being Basque, it was for being left-wing.”

    Ask a Basque nationalist.
    It was bombed for both reasons.

  49. @RogerH,

    My humblest apols. make that 191 votes per seat.

  50. @ greengrass
    “I have also seen how Zapatero in Spain is being cold-shouldered by Sarkozy, Merkl etc. (France Soir 14 March). An example of what happens to a lightweight leader.”

    Nonsense on two counts. Firstly, Spain, irrespective of its leader, is in the ‘second tier’ of EU importance (Germany, France UK and, for politeness sake, Italy, being the first). Second, Zapatero is not a lightweight, having won two elections against the odds and even now, with the worst unemployment rates in the EU, is still only three points behind in the polls. A situation which augurs well for him in 2012.

1 2 3