Impact on electoral administration

Impact on electoral administration

The Returning Officers and their staff who were responsible for the May 2018 elections successfully ran the pilot schemes in all of the five pilot areas:

  • Polling station staff told us that they had no difficulty checking people’s identification
  • They were confident that they could do this again at a future election
  • Across all areas, additional staffing and training were required for the pilot

Impact on electoral administration

These schemes were run at polls with a low turnout, with fewer people voting and showing identification in polling stations. As shown below there are clear lessons that can be learned for future polls where turnout is higher with more people voting in polling stations.

Delivering the identification requirement has specific implications for the administration of the poll and these are set out in more detail in our individual reports on each pilot scheme. We have considered the administrative impact across four aspects of delivery of the polls below.

Overall findings from the pilot schemes

All of the areas increased their staffing in order to deliver the pilot. There were varying approaches to this, ranging from employing ten extra polling station inspectors in Swindon to using one third more polling station staff in Bromley. Clearly there were also additional costs associated with increased staffing. For example, across the pilots the additional staff and training costs ranged from very little to a third of the usual costs at local elections.

However, in feedback after polling day the pilot areas told us that this increase in staffing would not be required to administer voter identification at future local elections. This is because additional pilot elements, such as the collection of data on polling day (required to inform the evaluation), would not need to take place. A few areas also did not think they would need additional staff at every polling day, even for a higher turnout poll such as a UK parliamentary general election.

More and/or longer training sessions were also required to support staff in delivering the pilot. However, feedback from electoral administrators in the pilot areas indicated that, while this training required planning and preparation, it did not present a significant challenge for them in running the pilot. We also know that the training was delivered well. Our survey of polling stations staff found that, on average, 94% rated the training they received as good or excellent. Close to 100% of polling staff also told us they were confident about the process they had to follow in order to check voters’ identification.

The 2018 pilots offered a limited test of the process and cost of issuing local identification to electors. Three pilots (Bromley, Gosport and Woking) offered a form of local identification but only Woking needed to issue any. This is likely to be because Woking had a narrower identification requirement where electors were slightly more likely to need to use a local option.

Woking issued 63 local elector cards and their feedback indicates that this process was manageable for them to deliver. However it did incur some additional costs associated with the production and delivery (where needed) of the cards. Gosport did raise a concern in their feedback about their ability to resource the local identification route if it had been significantly used. This was mainly because, as a small local authority, they would have limited flexibility in drawing on resources from other teams.

Swindon and Watford both used IT in the polling stations to scan QR codes on electors’ poll cards. In both pilots the systems worked well and there were no notable issues on polling day related to the IT. Also, the systems provided Returning Officers with useful, live information on turnout at polling stations that they would not normally have access to.

However, planning and setting up these IT systems required a significant amount of time and resource commitment from the electoral administration teams, the software suppliers and Cabinet Office. This commitment of time and resource stems largely from the level of security needed to run these systems, which hold significant personal data, as well as the level of assurance and resilience needed to ensure no problems arise on polling day.

While much of the work was associated with the development of the software, and would not necessarily be needed in the future, there would still be a sizeable level of commitment needed from software suppliers to support any local authorities sites using these systems at future polls. There would also be ongoing costs which local authorities would need to meet in order to use these systems. These include software licensing, hardware (tablets/scanners) and additional staff training.

Given the cost and time needed to support the use of these systems, the merits of including any IT- enabled pilots in future schemes needs to be weighed against the ability to scale up these systems across Great Britain. It 12 is also not clear from these pilot schemes that additional IT in polling stations (for example to scan barcodes or QR codes) is absolutely necessary to support the use of the poll card as a form of identification.

Feedback from Returning Officers and their staff indicate that delivering the identification requirement on 3 May did not present significant challenges. In response to our survey, 77% of polling station staff said they were very satisfied with how polling day went. This agrees with the feedback we received from Returning Officers which said that few issues had arisen on polling day and that they had received few or no negative responses from the public. These findings also agree with the observations made by Electoral Commission staff on polling day across the five areas. We saw few issues and largely observed processes working well.

A high proportion of staff (69%) also agreed with the statement that asking voters to prove their identity had little or no impact on their work on polling day. This also suggests that the variation in identification requirements did not make the task in polling stations notably easier or more difficult. Staff in Swindon and Watford (using poll card scanning) were more likely to agree there was little impact but there was little difference across the other areas. The difference in the quantity of different acceptable identification types between Bromley/Gosport and Woking do not, for example, appear to have made a significant difference to staff.

We also asked polling station staff if they would feel confident about replicating the requirement to show identification at a future poll and 97% said they would be confident in doing so.

Beyond the pilots: the impact on administration at future elections

The evidence we have gathered shows that the polls in the pilot areas on 3 May were well run and that the administrative challenges presented by the voter identification requirement were met by the Returning Officers and their staff. Looking beyond the pilots, there could be different administrative challenges which these pilots have not tested, including the need to process and deliver significant volumes of local electoral identification cards/letters.

The impact and risks for people running the elections could be different at elections with higher turnout with more people voting and showing identification in polling stations, such as a general election.

Turnout at the May 2018 elections, as is usually the case for local government elections, was relatively low, meaning that the pilots were run on a different scale than might be the case at a UK parliamentary general election. At these local elections between 20-30% of the electorate voted at a polling station whereas at a UK parliamentary election over 50% often vote at a polling station.

For example, staff in some polling stations in Watford said that scanning the poll card caused delays and queues which could be longer with more people voting. In areas such as Woking, which had a more limited choice of identification, more local elector cards may need to be processed and issued putting pressure on local authority election teams.

No applications were made in Bromley for the certificate of identity or in Gosport for the electoral identity letter, which means that the pilot did not test the resources required to process applications and issue certificates or letters.

The Returning Officers and their staff in Bromley and Gosport told us that if the list of acceptable identification was reduced they would expect the number of applications for certificates and letters to increase. They had some concerns about the pressure this could put on an already stretched elections team. This was particularly true for Gosport which, as a smaller authority, has less flexibility in the size of the elections team and their ability to draw on resources from elsewhere in the council.

One particular issue which the pilots in 2018 were ready to test was how to ensure privacy for voters who show photo identification but need to remove, for example, a head scarf to allow polling station staff to confirm their identity.

However, several of the pilot areas do not have significant British Asian populations. In Woking, which does have a significant Asian or British Asian population, no-one requested to show their identification in private. In Watford, where some wards also have a significant British Asian population, the availability of poll cards as part of the identification requirement meant that people may have been less likely to need to use this facility. Ensuring voters have the privacy they need could be a more significant administrative challenge in some areas and that should be a consideration for future testing.