My most recent post on this subject was constructively challenged by someone I respect highly. This person asked:

1. What are ‘high logistical and environmental costs’, in £££ and time?
2. How many people are affected by mobility issues, visual impairments or other things preventing them from going to polling places.
3. Concerning proxy voting, postal voting and statutory aids for people with visual impairments:
   1. How many people are involved?
   2. How many would benefit from the change?
4. How often does it happen that polling places become unavailable?
5. What would success in increasing turnout (due to e-voting) look like?
6. Concerning ‘facilitating reliable and rapid results (including fewer spoilt ballots)’:
   1. What speed up could be expected?
   2. How many spoiled ballots happen?
   3. And under e-voting, how many errors?
   4. What are the risks & consequences of hacking?
7. Concerning ‘reducing the costs of paper ballots’:
   1. What does it cost?
   2. How much would it cost to run e-voting instead?
8. Interesting question is how/whether e-voting helps or hinders a sense of community…

This post attempts some back-of-a-fag-packet answers to questions 1, 2 and 3. The other questions will have to wait for another post.

What are ‘high logistical and environmental costs’, in £££ and time?

To begin to answer this, I’ve had a quick look for costs of recent UK elections. Several sources agree that the 2017 general election cost around £140 million. The relevant Guardian article states

In a written statement, Chris Skidmore, a Cabinet Office minister, said £98m was spent by returning officers to run the poll and a further £42m on posted material.

A 2014 article on Democratic Audit UK has the headline The UK spends approximately £150 million per year administering elections. It states that the 2010 general election cost £113.3 million (£3.78 per vote). The Electoral Commission (EC) report Costs of the May 2011 referendum on the UK states This is the first time that a full report has been published on the costs of running a national poll. Among this reports ‘key facts and figures’ are

• The cost met from funds overseen by the UK Parliament of running the
  referendum was just over £75 million.
• 440 Counting Officers set up 42,800 polling stations
• 119,500 staff worked in those polling stations
• Over 80,500 staff worked on verifying and counting the ballot papers
• Nearly 7.2 million postal votes were issued to electors, of which 5.2 million (72%)
  were returned
• The EC distributed 27 million information booklets to
  households across the UK, reaching 96.1% of all households.
• The costs of the referendum met from funds overseen by the UK Parliament were
  just over £75 million. Other costs were, of course, incurred, most notably the spending by campaigners funded from donations and the costs to broadcasters of transmitting referendum campaign broadcasts.
• £58 million was paid to Counting Officers in fees and reimbursement of costs. £17
  million was paid out directly by the EC. This included
  • £8.5 million cost of campaign group mailings
  • £0.3 million postal vote ‘sweeps’ (searches of mail centres on polling day, to collect any postal ballot packs which might have been posted by electors on or very close to polling day, in order to deliver them to the Counting/Returning Officer before the close of poll.)
  • £7.5 million EC public awareness activity
  • £0.3 million EC grants to campaign groups
  • £0.1 million EC additional staffing
  • £0.5 million EC costs of administering payment of fees and costs to counting officers.

So let’s have a look at what might be saved by e-voting. I won’t count the campaign group mailings costs or grants to campaign groups because in elections, such costs would be borne by the parties and their candidates. Nor will I count the public awareness activity costs: I suspect these will remain similar (but with a different message ‘to e-vote, do this…’). Nor would broadcast costs count. So

• Let’s assume that the number of polling places remains the same, because we need to ensure that people who don’t have IT kit can vote. So the number of polling-place staff will remain the same. However, we will need to pay for them to be trained to deal with the new system. At a guess, £50 per trainee – so an extra £6 million.
• If polling places are local authority libraries and schools that have PCs in place, then costs of setting them up should be minimal. Otherwise, if such PCs have to be borrowed, moved and set up, let’s guess at £100 per polling place – so an extra £0.5 million.
• Let’s assume that only 500 staff are needed to run the electronic systems that count and verify the votes, and that this saves 8 hours per person. Let’s also assume that that these staff cost the same as I cost my employer – around £30 per hour. So we save 80,000 people * 8 hours per person * £30 per hour = £19.2 million saved. Hurrah!
• According to the table on page 73, postal votes cost £10.6 million. If all such costs are eliminated by e-voting, we save all of this. (Let’s assume that everyone who would have voted by post either pays for their own taxi to the polling place, buys their own IT kit or uses a volunteer digital proxy.)

So an all-electronic vote in 2010 might have saved up to around £35 million. If we double this because the 2017 general election cost around £140 million, we have potential savings of £70 million, and 640,000 staff-hours.

I have no facts which enable an estimate of how much time e-voting would save in calculating the result. Because the votes would be block-chain ‘transactions’, each individual vote should take a fraction of a second to register. Running checks and counting totals – maybe a couple of hours?

I would take the above with a very large pinch of salt. For example

• I’ve not included the cost of the hardware and software for processing the votes. 
• I’ve assumed all e-voting.
  • If there is a mixture of e-voting and paper voting at polling stations, we will need some of those £80,000 staff back. In fact, we may need more staff to add the e-votes in each constituency to the paper votes.
  • If we allow only e-voting and postal voting, the number of postal votes might even rise so people who cannot e-vote can still participate.

How many people are affected by mobility issues, visual impairments or other things preventing them from going to polling places.

Concerning proxy voting, postal voting and statutory aids for people with visual impairments:

• How many people are involved?

• How many would benefit from the change?

In the 2017 general election, according to Electoral Commission data,

• electorate 46,835,433 people
• proportion of postal voters was 18%
• rejected ballots 0.2%
• rejected postal votes 2.4%
• turnout 69%
• polling station turnout 66%
• postal vote turnout 86%

I presume that means that that the number of postal votes was 8.4 million. According to the RNIB, 2 million people in the UK live with sight loss. Using the overall turnout rate of 69%, 1.4 million of these people would have voted. So maybe 7 million people used postal votes for other reasons.

I’ve not found any data on the Electoral Commission website for numbers of proxy voters. According to a 2017 newspaper report, the UK average number of proxy voters is 0.32%. That would imply around 153,000 proxy votes in 2017.