Memorandum by Electoral Reform Services
(ERS) Ltd (POS 23)
EXECUTIVE SUMMARY
— The design of election material should
follow three basic principles:
— require the personalisation of as
few items as possible (preferably just one) so as to avoid "matching"
when enclosing; and
— avoid complicated finishing processes.
— Current proposals for postal voting
give cause for concern as to their scalability.
— Prominent amongst these concerns
is the requirement to only bar-code ballot papers and not print
a number on them and the use of an inner security envelope (envelope
"A") that is printed with the ballot paper number and
needs to be "matched" with the ballot paper.
— The design of the postal voting material
seems to be being influenced by the requirement to produce a daily
marked register. This requirement is achievable in a number of
different ways and should be left for local authorities to choose
the approach that best suits them.
— Using a window in the inner security
or having the ballot paper sealable in some way removes the requirement
for this item to be personalised and therefore "matched"
when enclosed. This will allow for just one item in the postal
voting pack to be personalised (a combined security statement
and ballot paper, separable by a perforation) and therefore reduce
the requirement for matching. This fundamentally reduces risk
and provides scalability.
INTRODUCTION
1. Electoral Reform Services is a supplier
of electoral services to the public and private sector. Our primary
service is postal balloting. ERS has been involved with the administration
of Local Authority Pilot elections since 2001. We have distributed
over 3.5 million ballot papers under the umbrella of the Representation
of the Peoples Act.
2. This submission to the Committee is concerned
only with the administration of all-postal voting. Matters relating
to electoral fraud, public perceptions, impact on turnout, access
and disability issues and voter choice are not covered in this
submission.
THE TIMETABLE
3. The current electoral timetable allows
for only twelve days to print and enclose the postal ballots and
associated documents, and have them ready for collection by the
Royal Mail or for delivery directly by the local authority. This
period runs from the close of nominations to the date for the
distribution of postal ballot papers (11th day before poll). For
any project this is a demanding timetable. When combined with
the complexity of the postal voting projects and the 100% accuracy
they require the timetable is challenging.
4. When deciding on the format of the election
material it is therefore necessary to consider the print and production
issues alongside the electoral issues.
DESIGN OF
ELECTION MATERIAL
Basic Principles
5. The design of election material for postal
voting should follow three basic principles.
Use common sizes
6. The greatest capacity and flexibility
within the print industry is achieved if common sizes are used
in the production process, ie A4 or A3. A4 or A3 size printed
items can be produced on a large number of machines, including
desk-top printers.
7. Envelopes and mailing machines are primarily
designed to handle these formats.
8. Use of non-standard sizes requires more
specialist printing machines and more complicated finishing processes.
9. Personalisation of non-standard size
printed items can also only be achieved on more specialist machines.
Personalise as few items as possible and avoid
"matching"
10. In a traditional postal voting pack
the various items are personalised with at least one piece of
common information, the ballot paper number. The separate items
are then collated and enclosed into a despatch envelope.
11. The more separate items that need to
be personalised the longer the production time and the greater
the complexity of the project. Further, if items do have to be
personalised then it is important that common sizes are used as
this again gives greater flexibility and control of the production
process. For example, personalising onto envelopes is a far from
straightforward process.
12. The challenge with the personalisation
of any printed item is not only making sure that all the items
are produced but also maintaining their running order. This is
so that when it comes to the enclosing process the separate items
are each available, in turn, at the correct time. Clearly with
a small number postal votes (ie not an all postal pilot) managing
this process can be easily achieved. However, with an all postal
election, with much larger quantities, control of the process
is more difficult.
13. To reduce the number of separate items
to be personalised two techniques have been utilised; joining
of items and use of windows in envelopes.
14. By joining items together production
time is reduced and when it comes to enclosing the items (filling
the envelopes) the requirement to match items together is reduced.
The solution that provides the lowest level of risk, greatest
capacity and quickest production time is to personalise one item
only. This is also the most cost efficient.
15. Windows on envelopes allow us to view
particular information enclosed within that envelope without opening
the envelope. For example, the address on a letter but not the
contents of the letter. The use of windows in envelopes contained
in the postal voting pack can reduce the number of items that
need to be personalised.
16. Each postal voter usually receives four
matched/personalised items. These are the ballot paper (numbered),
a security statement (numbered), an inner envelope (numbered)
and a bar-coded reply envelope. These items are ordinarily enclosed
into an addressed despatch envelope with some form of information
leaflet. Thus, five items need to be "matched" for enclosing
purposes. This operation can only be done manually. Manual filling
of "matched" items is a slow and costly process.
17. To achieve the deadlines/timetable in
an all postal ballot the process either has to be mechanised or
requires a large number of people to manually fill the election
material.
18. Filling matched items is also possible
on machines using intelligent mailing. This is where barcode/omr
readers positioned on the machines check the items being enclosed.
Unfortunately however, intelligent mailing does not work well
with more than two matched items. It is also slower than non-intelligent
mailing machines, more costly and there are fewer machines available
within the industry with the required technology.
19. Manual enclosing is always an option
for matching processes. But, to achieve the required throughput
usually requires a significant number of "bodies" to
complete the job.
20. Non-intelligent mailing does not (unless
run extremely slowly speeds) allow for the matching of any items.
However, non-intelligent mailing machines are common throughout
the mailing industry, have a higher production output and a lower
cost.
21. To give an indication of relative production
output:
— intelligent mailing machine = 2,000
envelopes/hour;
— manual enclosing (experienced) =
250 envelopes/hour; and
— non-intelligent mailing machine =
4,000 envelopes/hour.
22. Thus to enclose the envelopes for an
authority of 150,000 electors would take the following period
of time for a European Parliamentary Election only, ie no production
breaks for ward changes .
— intelligent mailing machines (assuming
two machines) = 37.5 hours; and
— manual enclosing (30 experienced
staff) = 20 hours.
23. Therefore, to enclose 1.5 million items
would require 200 hours manual enclosing (8.3 days, assuming 24
hour production) or 375 hours (15.6 days) intelligent mailing
enclosing. Clearly more staff and more machines could be utilised
but then scalability and control become a problem.
24. The difference in cost of enclosing
envelopes is approximately doubled each time the process moves
towards hand enclosing. The cost of manual enclosing = 2* intelligent
enclosing = 2* non-intelligent enclosing.
25. Regardless of production times, capacity
issues and cost, the fundamental reason why matching should be
avoided is risk. Each of these processes (manual or machine) requires
quality assurance and control mechanisms. However, to maintain
the integrity of enclosing processes that require matching needs
checking and auditing processes for each individual envelope and
this requires time. Time is not a commodity that is spare in the
current electoral timetable. It may seem an easy statement to
make that it seems relatively straightforward for someone to get
the correct items into an envelope. However, that may be the case
when dealing with small quantities but when dealing with all postal
voting it is a complete differently situation.
26. There is no other mailing that we can
think of where the incorrect enclosing of two items could throw
such doubt on the process that the entire project is placed in
doubt. For example, in a bank statement mailing if two individuals
receive each others statements then whilst there is concern from
those two individuals and the banks reputation suffers in the
minds of those two individuals, everybody else in the mailing
is totally unaffected. In an election mailing this is not the
case. Thus, every effort should be made to avoid the possibility
of the incorrect items being sent to any individuals and the most
effective way of ensuring this to avoid matching items by using
the methods described above, namely joining items and window envelopes.
Avoid complicated finishing processes
27. Intricate folding, gluing, tip-on's
(glued on sections), scratch-off panels all add to the length
of time an item takes to produce. They also add cost and increase
risk. Essentially the more processing an item has to go through
before it is finished the greater the number of spoils.
28. With regard to scalability the more
complicated the finished item the fewer machines there will be
available to produce that item.
THE CURRENT
THINKING
29. We have not as yet had the opportunity
to view formally the proposals for the forthcoming electoral pilots
in the European Parliamentary Regions. However there are a number
of proposals that have been circulated that give us cause for
concern.
Bar-coding of Ballot Papers
30. Bar-coding of ballot papers requires
specific personalisation using different machines to those that
traditionally print ballot papers. This in itself may not be an
issue. However, our understanding is that the proposal is for
the bar-code alone to be printed on the ballot paper and not the
ballot paper number. This proposal would create significant problems
for the production and enclosing process. As part of the enclosing
process it is necessary for operators, using either machines or
manual processes, to be able to view and check the items being
enclosed. Barcodes cannot be read visually.
31. Barcodes also occasionally fail to be
able to read (and voters tend to scribble over them). How will
the ballot papers be checked if this is the case?
32. Only bar-coding the ballot paper increases
risk. The ballot paper number should also appear on the ballot
paper.
33. Bar-codes are required for the purpose
of electronic counting. They may also assist with the matching
of un-matched ballot papers and security statements. However,
they are far from essential to a process that can easily be achieved
by placing the items in number order and inputting them into a
database.
Window Reply Envelope and Numbered Inner Security
Envelope
34. Some pilots in 2003 used window reply
envelopes.
35. Why use a window reply envelope? Firstly,
as it is a non-personalised item, it reduces the requirement for
it to be "matched" when filling the despatch envelope.
Avoiding matching and personalisation as detailed above is a good
thing.
36. To return his ballot paper the voter
is required to place one of the items within the electoral pack
(usually the security statement) into the reply envelope in such
a way that the return address of the authority shows through the
window of the reply envelope as well as a personalised bar-code
specific to that voter. This enables the Electoral Registration
Officer, without opening the envelopes, to create a list of individuals
who have returned envelopes and produce a marked register.
37. Within the reply envelope the voter
should have placed his ballot paper in a numbered security envelope
(envelope "A") so that when the envelope is opened and
the security statement/declaration of identity is checked his
ballot paper is safely sealed in the security envelopes and his
voting intentions hidden. Unfortunately the use of the inner envelope
still requires matching.
38. Some authorities using the window return
envelope also avoided the use of the inner envelope by either
just not using it or by having sealable ballot papers (ie ballot
papers that can be folded in half and glued shut). These authorities
applied the basic principle of avoiding matching with regard to
the security envelope. The use of glue however can be problematic.
They also combined the security statement and ballot paper together
so that the net-result was a mailing pack that contained no items
that need to be matched, i.e. a combined ballot paper and security
statement, window return envelope and leaflet—and no inner
envelope.
39. It is however, in our view, preferable
for the voter to seal his ballot paper in a security envelope.
How then would it be possible to remove the requirement for this
to be personalised? This could be achieved by making the inner
envelope a window envelope into which the ballot paper is sealed
but through which the ballot paper number could be viewed.
40. Therefore, if we are able to combine
the ballot paper and security statement into one item separable
by a perforation (with a section being used as the address carrier),
have a non-personalised security envelope with a window, a return
envelope with a window we would have a mailing pack that required
no matched items.
41. This approach would require the voter
to put two items into envelopes in such a way that barcodes or
numbers were visible through windows. This could result in the
voting process becoming too complicated and therefore prohibitive.
Unfortunately, some voters are not able to follow the process
and put the items in the envelopes in the wrong way. If the item
to show through the window in the reply envelope is placed incorrectly
this could result in the ballot envelope not being returned to
the local authority.
42. The primary purpose of the window reply
envelope is to be able to produce a daily marked register without
opening the reply envelopes. It is recognised that this approach
will only create a list of those that have returned envelopes.
This list will not necessarily therefore be those who have voted
as the contents of the envelope is unknown. Provided all parties
accept this fact then creation of a marked register in this way
is acceptable.
43. The bar-coding of the envelopes by the
local authority takes time and for those authorities who will
have to open on a daily basis (to meet the demands of the election)
it is an unnecessary extra handling of the reply envelopes. It
seems over complicated to handle the envelopes once for one purpose
only to handle them again for another purpose within a short space
of time. Why not simply just open the envelopes, complete the
verification process and then when the ballot papers are secured
in ballot boxes bar-code read the security statement. Thus creating
a more accurate marked register. This would allow us to remove
the requirement for a window reply envelope, leaving the voter
with perhaps just an inner security window envelope. The reply
envelope would simply be envelope addressed to the local authority.
This would also reduce the risk of items not arriving.
44. The window reply envelope may be appropriate
for smaller local authorities who are not required to open envelopes
on a daily basis. For this reason the requirement of a daily marked
register should be left to local authorities to achieve in the
manner that best suits them rather than in a prescribed way.
45. An authority with 200,000 electors on
a 40% turnout would receive 80,000 return envelopes, if evenly
distributed over the voting period (10 days) this would be 8,000
a day. This quantity requires daily openings and verifications
of ballot papers.
46. The inner security window envelope still
may be confusing for voters but it is far less fundamental if
they fail to place the ballot paper in the inner envelope correctly
than if the reply envelope fails to arrive altogether.
47. The current proposal seeks to address
one issue, the creation of a marked register, but over complicates
another, the printing and distribution of ballot papers. The success
of the latter is far more fundamental to the success of the postal
voting process and it is our belief that everything should be
done to reduce the risk of errors occurring.
48. A postal voting pack should therefore
be designed that requires only one piece of personalisation and
therefore no matching. This requires various items to be combined,
the security statement and ballot paper(s) and the address carrier.
It should also incorporate either a window security envelope or
some manner in which the ballot paper(s) could be sealed and a
non-window reply envelope.
49. A postal voting pack produced in this
manner will make available to local authorities a significant
section of the printing and more importantly distribution industry.
This will bring all the cost efficiencies that come through a
competitive market.
CONCLUSION
50. ERS's experience of conducting all postal
ballots enables us to make the above observations concerning the
administration of postal voting projects. We are concerned that
the current proposals for postal voting could result in a format
that may be achievable for limited postal voting projects but
when scaled to the demands of an all-region pilot are not cost
effective, not efficient and add considerable risk to the process.
| 
