31

In the UK on Thursday 4 May 2023, local elections are taking place across England. To vote at a polling station, you will need to bring photo ID. This is a new law introduced by the current government. It may be worth noting that the current government voting base is heavily skewed towards the older population and it includes a provision that the Oyster Card only counts for over 60s.

Is there anything put in place to count those turned away because of missing identification? What are we likely to know after the election about the true effect of this change in the law?

User65535
  • 3,552
  • 1
  • 14
  • 30
  • 3
    I presume you mean to ask if we will ever know the number of otherwise eligible voters who are turned away. – EvilSnack Apr 27 '23 at 03:02
  • 1
    I get the suspicion that the core idea in the title is inherently relying on the absence of evidence being evidence of absence, which is not a reliable conclusion to draw. – Flater Apr 27 '23 at 04:09
  • 1
    I'm sure there will be academic studies because it's an important topic; and maybe the Electoral Commission or other bodies will also investigate. Doubtless they will use a combination of surveys, public data, and statistical analysis. But I can't provide specific information on these studies. (Note that there is no way of recording those who stay home, or who would have voted but heard about the requirements too late to take action, other than by survey or by indirect methods such as analysing turnouts.) – Stuart F Apr 27 '23 at 08:29
  • Summarizing the answers below: yes to the 1st Q, and perhaps no to the 2nd, LOL. Of course it depends how you define "turned away". – the gods from engineering Apr 28 '23 at 00:36
  • You have asked two different questions: how many will be "denied a vote" and how many will be "turned away". Any measurement of the effect is going to depend very strongly on exactly what you choose to measure. You have also asked the question in two different ways: "will we know?" and "is there anything put in place to count?". Again, those are different questions. – Michael Kay Apr 28 '23 at 08:21
  • 3
    I'm not sure what the sentence "It may be worth noting..." adds to the question: it seems to suggest you are also interested in the answers to other questions, such as the age profile of those who may be "denied a vote". That's certainly an important question (most studies indicate that the old are more likely to lack suitable ID than the young) but it's out of scope for the question you have actually asked. – Michael Kay Apr 28 '23 at 08:27
  • 9
    I was curious why the Oyster Card thing would be restricted to 60+ but there's a pretty good reason: If you are over 60 the Oyster Card is a government-issued photo ID for free transit rides. None of the other Oyster cards require verification to obtain, nor do they have a photo on them. I'm not sure why you included it in your question. – Machavity Apr 28 '23 at 13:06
  • 1
    I hadn't noticed the student Oyster card was also a photo ID, but the list of requirements is far less stringent. Basically anyone who is a student in London (UK citizen or not) can get one. – Machavity Apr 28 '23 at 13:19
  • 1
    The point I am trying to make it that the ruling party has come up with a set of rules that includes ID that most older people have but not the ID that younger people have, and that is what would be expected if they want to have more older people voting. If they were more interested in having a vote that represents the true wishes of the population they would ensure we know the effect of this rule change. This is relevant to the motivation for the point of the question. – User65535 Apr 28 '23 at 13:25
  • 3
    There's an assumption in this question that people turned away don't just mutter "oh bother", then go home to collect some ID, and return. (For foreign readers, except in the most rural of locations one's polling station is unlikely to be more than one km from residence, and it's open from 6am to 10pm). – nigel222 Apr 28 '23 at 13:55
  • 3
    @Machavity The fact that non-citizens and under 18s can get a student Oyster card is no reason to exclude them, the same is true for many of the other IDs allowed. A person still needs to be eligible to vote once they show up. – Princess Ada Apr 28 '23 at 14:52
  • i suggest that's not greatly likely and either way, it looks more than anything like another thinly-disguised attempt to groom us into accepting identity cards.

    How stupid do you think our rulers believe you are?

    Who previously failed to notice the expensive and otherwise wholly pointless change from simple 'Passport…' to 'Identity and Passport Service…'?

     – Robbie Goodwin Apr 30 '23 at 18:25
  • I suppose one could extrapolate from Canadian data as Canada also, for some time, required voter id. – paulj May 02 '23 at 17:55
  • Further to the 'Identity and Passport…' name change and as if to prove the value of the whole exercise, my sister was asked to remove her sunglasses, which could hardly have had less of a tint and certainly did not make her harder to recognise. I 'Identified' myself using a cancelled driving licence from an address two counties and 80 miles away, which I left in 2008, and bearing a 20-year-old photo with no hint of the beard I have now. – Robbie Goodwin May 15 '23 at 18:46

2 Answers2

36

No. The number of people denied a vote, counted and recorded by the new legislation, as described in CDJB's answer, can only be considered a minimum possible number of those denied their chance to vote.

Perhaps most importantly during an Urgent Question placed to the government on Thursday. The electoral commission have confirmed that voters turned away from polling stations by 'greeters' will not be counted as having been unable to vote.

A minister has repeatedly refused to tell MPs whether the number of voters turned away from polling stations because of a lack of ID at local elections next week will be properly recorded, arguing that it is ultimately up to individuals whether or why they decide to vote.

At the end of the debate, Betts used a point of order to say the Electoral Commission had informed him that greeters would not count the numbers of those they turned away.

The actual number additionally will include;

  • Those who don't show up because they have no id at all
  • Those whose newly requested id does not arrive in time or whose id has been held up by some other bureaucratic process (renewal or using id a to get id b
  • Those who can't find their id on the day so don't leave the house
  • Those who find looking for their id to be just one more reason not to bother.

And of course the collection of this data is a new system and we have no way to estimate how rigorous or accurate it will be at this time.

Jontia
  • 24,192
  • 4
  • 94
  • 120
  • 7
    We can make a fair stab at an estimate by looking at the turnout for prior elections, then comparing the turnout for this one. – Valorum Apr 26 '23 at 20:21
  • 24
    @Valorum Local election turnout has over the last couple of decades been between 30% and 40% with the odd 60% thrown in. I don't think there is enough stability in the numbers for any such inference. https://commonslibrary.parliament.uk/research-briefings/cbp-8060/ – Jontia Apr 26 '23 at 20:39
  • Comments discussing how voter suppression could be estimated by looking at election turnout have been moved to chat; please do not continue the discussion here. Before posting a comment below this one, please review the purposes of comments. Comments that do not request clarification or suggest improvements usually belong as an answer, on [meta], or in [chat]. Comments continuing discussion may be removed. – Philipp Apr 27 '23 at 13:09
  • @Jontia I think it's a little flawed to compare the turnouts of local elections held on their own with those held in parallel with a GE (2010, 2015, etc)? – CDJB May 05 '23 at 07:09
  • @CDJB you're right. And the 60s do correlate to the GEs. But even 30% to 40% represents a variation of 1/4. That's too swingy to assign any change this year to a specific cause. – Jontia May 05 '23 at 14:04
21

Yes, paragraph 7 of Schedule 1 of the Elections Act 2022 amended the Representation of the People Act 1983 to include a requirement for the Secretary of State to publish a report on the effect of voter ID on the May 2023 local elections and the next two parliamentary elections.

It also, in paragraph 24, amended the elections rules to introduce a duty for the various officers involved in the election - the presiding officer, the returning officer, and the registration officer - to collect and anonymise prescribed information relating to applications for ballot papers made under the new voter ID rules in order to inform the Secretary of State's report.

Furthermore, the Electoral Commission has undertaken to conduct public opinion research to evaluate the effect on those who do not attend a polling station to be counted, according to Cat Smith, a member of the Speaker's Committee on the Electoral Commission:

Jeff Smith: It is good news that data will be collected on the number of voters who get turned away for not having ID at a polling station, but as we all know, parties often have tellers outside who will remind people about the voter ID requirements, so how can the Electoral Commission collect data on voters who turn away before they even get into the polling station?

Cat Smith: My hon. Friend, who is a seasoned campaigner and is familiar with the scenes outside polling stations, has identified the potential gap in the data. Of course, polling station staff will not be able to collect data from people who do not go into the polling station. However, the commission has identified that as a potential issue and will undertake public opinion research on the reasons why people did or did not vote in the elections.

So yes, data on those affected by the new rules will be collected, and the information published after the May 2023 local elections, as well as the next two parliamentary elections. This will presumably take a similar form to the government's report on the 2019 voter ID pilot.

CDJB
  • 106,388
  • 31
  • 455
  • 516
  • 9
    Unless the report prepared includes things like survey data to estimate how many people didn't even bother due to the new requirement, I have to agree with Jonita's answer that this would only be a lower bound estimate. An upper bound estimate would also be possible by comparing turnout rates in this election to turnout rates in previous elections and assuming that past record high turnouts would represent an upper bound. An exact number, of course, is probably unknowable. The methods of the 2019 report were pretty iffy in some respects. – ohwilleke Apr 26 '23 at 22:43
  • @ohwilleke "An upper bound estimate would also be possible by comparing turnout rates in this election to turnout rates in previous elections and assuming that past record high turnouts would represent an upper bound." The turnout rates are public information I guess, so probably someone in academia will do that and that upper bound will likely also be available. – NoDataDumpNoContribution Apr 27 '23 at 07:18
  • 5
    That upper bound will be useless, because turnout in local elections is extremely variable ─ according to here, average local election turnout was 63.5% in 2010 and then 31.3% just two years later. So the upper bound might say that up to ~30% of all voters could have been prevented from voting, which would not narrow it down at all. – kaya3 Apr 27 '23 at 15:52
  • Notably, the 2019 report doesn't do a difference within difference methodology analysis which is one of about three viable approaches which taken together might provide a credible answer. See, e.g., https://www.publichealth.columbia.edu/research/population-health-methods/difference-difference-estimation – ohwilleke Apr 27 '23 at 19:54
  • @kaya3 It's not that variable - in 2010 the locals were held simultaneously with the general election, same in 2015. For local elections on their own 30-odd percent turnout is about right. – CDJB May 05 '23 at 07:06
  • @CDJB Even if you look at only years with no general election, the range is between 47.5% in 1991 and 28.8% in 1998, a spread of about 19 points. The upper bound is still useless ─ if as many as 19% of eligible voters would be prevented from voting because of the ID rules, we'd know about it not from statistics collected from polling locations, but because there would be riots. – kaya3 May 05 '23 at 16:00
  • 1
    @kaya3 I agree, I just think the broader problem is that comparing any local election turnout to another is fundamentally flawed because of the different electoral circumstances; these local elections are in completely different areas than last year, for example. In 1991 the unitary councils contested yesterday didn't even exist. I agree that comparisons of turnout figures would not be sufficient to evaluate the effect of this policy though. – CDJB May 05 '23 at 16:31