A Closer Analysis of Postal Voting in Tower Hamlets

Posted on | Wednesday, 2 May 2012 | No Comments

I have previously blogged that 41% of votes cast in the recent Spitalfields and Banglatown bi-election were postal. This staggeringly high figure included all postal votes received, including those subsequently rejected. The reason we need to focus on total postal votes received as a percentage (of all votes cast), rather than the lesser percentage after correcting for rejects, is because the rejected figure is a subjective one. To explain, received postal votes and enclosures are checked and counted numerous times under dual control to verify the number of postal votes received - 956 in this case - were correct.  However, the 'rejected' count on the other hand, is arrived at through a process of subjective judgements, not necessarily conducted under dual control, and that is why the figure of 135 rejects is unreliable. As I'll explain, the already high reject rate of 14% (135/956) could well have been considerably higher if for instance a more stringent, professional judgement had been made on signature irregularities. Alarmed yet? You should be, because this raises serious doubts over the validity of Gullam Robbani's marginal win by only 43 votes.

I'm not unfamiliar with either the business process or technology of voter registration, including voting and signature recognition/verification systems. And that's why I wanted to see first hand, as an observer, how Tower Hamlets were managing the postal vote processing for the forthcoming 3 May elections. My observations also provide an insight into the reliability of the Spitalfields and Banglatown bi-election (reject) figures above.

Over a 2 day period I randomly observed the processing of 20 batches each of 50 Postal Vote Statements (PVS) at Tower Hamlets Town Hall. To explain, a PVS accompanies each ballot paper and is used to verify the authenticity of each postal vote by comparing address, date of birth and signature against details held on the electoral register. Each batch of 50 PVS's is fed into an automated recognition system and irregularities are highlighted on screen to be accepted or rejected by a council official. Approximately 50% of PVS result in some form of mismatch against the electoral register (47% in the sample I observed). Of these mismatches, approximately 10% are subsequently rejected by the council official (7% in the sample I observed). In summary, of the 1,000 PVS I observed, 35 or 3.5% were rejected. Significantly lower than the figures produced for the Spitalfields and Banglatown bi-election. I should add that in my opinion, a higher  percentage of signatures appeared fraudulent. The implication being that the Spitalfields figure of 14% was probably also too low. 

In conclusion, in light of the exceptionally high and unprecedented postal vote and rejection rate during the Spitalfields and Banglatown bi-election, there is ample justification for the Metropolitan Police to:

1. Interview each of the 956 constituents who submitted a postal vote to establish whether any were forced to vote under duress (a clear risk in postal voting). 

2. Examine each of the 135 rejected PVS and in the case of fraudulent signatures/submissions, prosecute as appropriate.

I should add that I have now written to Tower Hamlets' Metropolitan Police Commander twice on this issue but have so far received no response. None of our 2 local parliamentarians have shown the courage to comment on the issue.

Unless robust Police action is taken, and seen to have been taken, the very basis of our democratic electoral system will continue to be undermined.


Leave a Reply

Search This Blog


Grenville Mills